On the radial abundance gradients in disks of irregular galaxies
aa r X i v : . [ a s t r o - ph . GA ] M a y Mon. Not. R. Astron. Soc. , 000–000 (0000) Printed 16 July 2018 (MN L A TEX style file v2.2)
On the radial abundance gradients in disks of irregular galaxies
L.S. Pilyugin , , , E.K. Grebel , I.A. Zinchenko , Astronomisches Rechen-Institut, Zentrum f¨ur Astronomie der Universit¨at Heidelberg, M¨onchhofstr. 12–14, 69120 Heidelberg, Germany Main Astronomical Observatory of National Academy of Sciences of Ukraine, 27 Zabolotnogo str., 03680 Kiev, Ukraine Kazan Federal University, 18 Kremlyovskaya St., 420008, Kazan. Russian Federation
Accepted 2015 April 22. Received 2015 April 22; in original form 2015 January 01
ABSTRACT
We determine the radial abundance distributions across the disks of fourteen irregular galax-ies of the types Sm and Im (morphological T types T = T =
10) as traced by their H ii regions. The oxygen and nitrogen abundances in H ii regions are estimated through the T e method or / and with the counterpart method ( C method). Moreover, we examine the corre-spondence between the radial abundance gradient and the surface brightness profile. We findthat irregular galaxies with a flat inner profile (flat or outwardly increasing surface brightnessin the central region) show shallow (if any) radial abundance gradients. On the other hand,irregular galaxies with a steep inner profile (with or without a bulge or central star cluster)usually show rather steep radial abundance gradients. This is in contrast to the widely heldbelief that irregular galaxies do not usually show a radial abundance gradient. Key words: galaxies: irregular – galaxies: abundances – ISM: abundances – H ii regions –galaxies: photometry The radial distribution of gas-phase oxygen abundances tracedby H ii regions has been investigated in the disks of many spi-ral galaxies (Vila-Costas & Edmunds 1992; Zaritsky et al. 1994;van Zee et al. 1998; Pilyugin et al. 2004; Moustakas et al. 2010;Gusev et al. 2012; Pilyugin et al. 2014a; S´anchez et al. 2014). Itwas found that almost all spiral galaxies show radial abundancegradients in the sense that their inner H ii regions (i.e., those closerto the galactic centers) have higher oxygen abundances than theouter ones.The radial distribution of abundances across the disks of irreg-ular galaxies is less well studied. Pagel et al. (1978) analyzed spec-tra of a number of H ii regions in the Small and Large MagellanicClouds. They determined the abundances in H ii regions through thee ff ective temperature ( T e ) method using their own measurementstogether with the spectral measurements by other authors and ex-amined the spatial distributions of abundances in those galaxies.Pagel et al. (1978) concluded that any radial abundance gradient inpresent-day abundances is small or absent in the Large MagellanicCloud and is conspicuously absent in the Small Magellanic Cloud.In the Small Magellanic Cloud, also stellar metallicity determina-tions support the absence of a radial gradient, although there is alarge metallicity spread of ∼ . / H] for a given age(Glatt et al. 2008; Cignoni et al. 2013).Roy et al. (1996) studied the oxygen abundance distributionsin the disks of the dwarf irregular galaxy NGC 2366 and thedwarf Seyfert I galaxy NGC 4395 using imaging spectrophotome-try with narrow-band filters in the lines of H α , H β , [O iii ] λ ii ] λ iii ] / [N ii ] as an abundance indicator (O3N2 calibration). They found that there is no globaloxygen abundance gradient across the disks of those galaxies.Hunter & Ho ff man (1999) obtained emission-line long-slitspectra of 189 H ii regions in a sample of 65 Im , Sm , and bluecompact dwarf galaxies. They estimated the oxygen abundancesin H ii regions using the line ratio [O iii ] / [N ii ] (O3N2 calibration)and the combination of R = [O iii ] + [O ii ] and [O iii ] / [O ii ] (two-dimensional R calibration) when the oxygen line [O ii ] λ ff man (1999) examined the radial abun-dance distribution in disks of eight Sm and Im galaxies for whichthey measured at least three H ii regions. They found that the oxy-gen abundances within a given galaxy generally vary by about 0.2dex, but they did not detect a trend in oxygen abundances with ra-dius except for the Sm galaxy DDO 204.Kniazev et al. (2005) measured oxygen abundances with thedirect method in three H ii regions in each of the dwarf irregularsSextans A and Sextans B. While they found Sex A to be chemicallyhomogeneous, one of the three H ii regions in Sex B turned out tobe about twice as metal-rich than the other two, and the abundancesof other heavy elements suggest an enrichment by a factor of ∼ . ii regions. Kniazev et al. (2005)attribute this to inhomogeneous chemical enrichment.van Zee & Haynes (2006) carried out long-slit spectroscopyof 67 H ii regions in 21 dwarf irregular galaxies. Oxygen abun-dances for 25 H ii regions were derived through the direct T e method; the abundances in other H ii regions were estimated usingstrong line calibrations. van Zee & Haynes (2006) considered theoxygen abundances as a function of radius for 12 irregular galaxieswith three or more observations and found that the abundances arevery similar (within the formal errors) within each galaxy with the c (cid:13) L.S. Pilyugin et al. possible exception of the galaxy UGC 12894. van Zee & Haynes(2006) noted that the radial trend in oxygen abundances (threepoints) in the UGC 12894 may be artificial because the abundancesof the inner and outer H ii regions were obtained via di ff erent meth-ods (through the strong line calibration for two inner H ii regionsand through the T e method for outer H ii region).Similarly, Lee et al. (2007) obtained oxygen abundances for35 H ii regions in eight dwarf galaxies in the Centaurus A group andin 13 H ii regions in closer dwarfs. Some of their measurements usethe direct T e method, while the majority of the abundance deter-minations is based on strong-line calibrations. Although the resultsfor individual H ii regions in a given galaxy tend to vary, Lee etal. point out that the variations are within the uncertainties of thestrong-line method. In one of the dwarf irregulars of the Cen Agroup, AM 1318 − ii regions is considerably moreoxygen-rich than the others. Lee et al. argue that the measured lineintensity ratios suggest that this emission nebula is a supernovaremnant. They also note that radial gradients may exist in someof their targets such as in the Sm NGC 3109 or NGC 5264, but thatmore and deeper data are needed to establish this.It is the current belief that irregular galaxies generally do notshow radial abundance gradients in their young populations and arechemically homogeneous. This implies that there is a “spiral versusirregular dichotomy” in the sense that there is a sudden change fromspiral (radial abundance gradients are usually present) to irregulargalaxies (typically no gradients). However, other properties (e.g.,gas fraction, global metallicity) vary smoothly in transition fromspirals to irregulars (Zaritsky et al. 1994; Pilyugin & Ferrini 2000;Garnett 2002; Pilyugin et al. 2007, among others).The measurements of the abundance gradients in the disks ofirregular galaxies often encounter the following di ffi culty. Reliableoxygen abundances in a number of H ii regions in the disk of agalaxy should be determined in order to evaluate the existence ofan abundance gradient. Abundance determinations using the direct T e method require high-precision spectroscopy including the weakauroral lines [O iii ] λ / and [N ii ] λ ii regions in a given irregular galaxy.The oxygen abundances in the other H ii regions are then estimatedusing the strong-line method pionered by Pagel et al. (1979) andAlloin et al. (1979). The principal idea of the strong-line method isto establish the relation between the (oxygen) abundance in an H ii region and some combination of the intensities of strong emissionlines in its spectrum (such a relation is usually called a “calibra-tion”). Di ff erent calibrations were suggested. A prominent charac-teristic of the calibrations is that they are not applicable across thewhole range of metallicities of H ii regions but only within a lim-ited interval (usually only at high or at low metallicities). The oxy-gen abundances of irregular galaxies typically lie within or near thetransition zone in the R – O / H diagram (from 12 + log(O / H) ∼ ∼ ii regions, which can be used overthe whole range of metallicities of H ii regions and which pro-vides oxygen and nitrogen abundances on the same metallicityscale as the classic T e method (Pilyugin et al. 2012, 2013). Us-ing this method, we examined the abundance gradients in thedisks of 130 late-type galaxies including several irregular galax-ies (Pilyugin et al. 2014a). In that study, radial abundance gradi-ents were found in irregular galaxies. Here we will focus on theinvestigation of the abundance gradients in a sample of irregular galaxies ( Sm and Im , morphological T types 9 and 10). Since thereis a relation between oxygen abundance and disk surface brightnessin spiral galaxies (e.g., Pilyugin et al. 2014b), we will also exam-ine the relation between radial abundance distributions and surfacebrightness profiles of the disks of irregular galaxies.The paper is structured as follows. The spectral and photomet-ric data are reported in Section 2. The radial abundance gradientsare determined in Section 3. The discussion and conclusions aregiven in Section 4, followed by a summary (Section 5). We have selected a sample of irregular Sm and Im and galaxieswith morphological T types of 9 and 10 according to the RC3 cat-alog (de Vaucouleurs et al. 1991). It should be noted that the mor-phological classification of some galaxies is not robust. The mor-phological T types in di ff erent sources can di ff er by up to 1. Weonly consider irregular galaxies with available spectra for four andmore H ii regions. The validity of the radial abundance is definednot only by the quantity and quality of the spectra but also by thedistribution of the measured H ii regions along the galactic radius.We reject galaxies where the measured H ii regions cover less than ∼ / ii regions cover onlya small fraction of the optical radius of the galaxy, which preventsa reliable investigation of a radial abundance gradient.Our final list includes fourteen irregular galaxies with opticalradii of R & T ) from the RC3 is reported in column 5.The right ascension (R.A.) and declination (Dec.) (J2000.0) of eachgalaxy are given in columns 6 and 7. The right ascension and decli-nation are obtained from our photometry (see Section 2.3) or takenfrom the NASA / IPAC Extragalactic Database ( ned ) . The positionangle (P.A.), axis ratio ( b / a ), and inclination are listed in columns 8– 10. The isophotal radius R in arcmin and in kpc of each galaxyis reported in columns 11 and 12, respectively. The adopted dis-tance d taken from Karachentsev et al. (2013) or from the ned isreported in column 13. The ned distances use flow corrections forVirgo, the Great Attractor, and Shapley Supercluster infall. The ref-erences to sources for geometrical parameters (first reference) andfor distances (second reference) are given in column 14. ii region spectra We use the emission line intensities in published spectra of H ii regions from di ff erent works for abundance determinations. Wehave searched for spectra of H ii regions with measured H α , H β , The NASA / IPAC Extragalactic Database ( ned ) is operated by theJet Populsion Laboratory, California Institute of Technology, undercontract with the National Aeronautics and Space Administration. http://ned.ipac.caltech.edu/ c (cid:13) , 000–000 bundance gradients in irregulars Table 1.
The adopted properties of our galaxies. n Name T R.A. Dec. P.A. b / a Inclination R R d ReferenceNGC UGC Other type degree arcmin kpc Mpc1 2023 DDO 25 10 02:33:18.20 33:29:28.0 1.00 0 0.83 2.24 9.30 RC3; K13 a ned ned ned
11 4449 7592 10 12:28:11.01 44:05:38.1 48 0.56 56 3.07 3.76 4.21 Here; K1312 CGCG 071-090 10 12:58:52.80 13:09:08.8 172 0.60 53 0.49 1.98 13.90 Here; ned
13 9614 10 14:56:47.70 09:30:33.4 21 0.77 40 0.50 7.23 49.70 Here; ned
14 12709 DDO 219 9 23:37:24.05 00:23:30.7 150 0.73 43 0.73 7.86 37.00 Here; ned a K13 – Karachentsev et al. (2013) µ (mag arcsec -2 ) R arcsec
NGC 2537 µ (mag arcsec -2 ) R arcsec
NGC 4214 µ (mag arcsec -2 ) R arcsec
NGC 4395 µ (mag arcsec -2 ) R arcsec
NGC 4449
Figure 1.
Comparison between the measured surface brightness profiles inthe SDSS r band (line) obtained here and profiles in the R band (open cir-cles) from Swaters & Balcells (2002). The arrow marks the optical isopho-tal radius R . [O iii ] λ ii ] λ ii ] λ + λ ii ] λ + λ iii ] λ We constructed radial surface brightness profiles in the infrared W µ m) using the publiclyavailable photometric maps obtained in the framework of the Wide-field Infrared Survey Explorer (WISE) project (Wright et al. 2010).The conversion of the photometric map into the surface brightnessprofile is discussed in Pilyugin et al. (2014b). Parameters such asthe galaxy center, the position angle of the major axis, and the axisratio are obtained through fitting of the isophotes by ellipses.We also constructed radial surface brightness profiles in the SDSS g and r bands using the photometric maps of SDSS datarelease 9 (Ahn et al. 2012). To estimate the optical isophotal radius R of a galaxy, the surface brightnesses in the SDSS filters g and r were converted to B -band brightnesses, and the AB magnitudeswere reduced to the Vega photometric system using the conversionrelations and solar magnitudes of Blanton & Roweis (2007).Swaters & Balcells (2002) reported surface brightness profilesin the R band for a large number of galaxies. Fig. 1 shows the com-parison between our measured surface brightness profiles in theSDSS r band (solid line) and R -band profiles (open circles) fromSwaters & Balcells (2002). Our surface brightness profiles withinthe optical isophotal radius R agree satisfactorily well with thoseof Swaters & Balcells (2002).All surface brightness measurements were correctedfor Galactic foreground extinction using the A V values fromthe recalibration of the maps of Schlegel et al. (1998) bySchlafly & Finkbeiner (2011) and the extinction curve ofCardelli et al. (1989), assuming a ratio of total to selective ex-tinction of R V = A V / E B − V = A V values given in theNASA Extragalactic Database ( ned ) were adopted. To transformthe surface brightness measurements to solar units, we used themagnitude of the Sun in the W V band and from its color ( V − W ⊙ = g and r bands were used to es-timate the isophotal R radius of each galaxy. The obtained radialprofiles were reduced to a face-on galaxy orientation. Note that theinclination correction is purely geometrical, and it does not includeany correction for inclination-dependent internal obscuration. Thevalues of the optical radius R determined here are listed in Ta-ble 1. There are no SDSS photometric maps for several galaxies ofour sample. The optical radii R (as well as the position angle ofthe major axis and the inclination angle) for those galaxies weretaken from the RC3 (de Vaucouleurs et al. 1991).The observed surface brightness profile of an irregulargalaxy can be fitted by an exponential (Swaters & Balcells 2002;Herrmann et al. 2013). There are bulges, bars, or nuclear star clus-ters at the centres of some irregular galaxies. A bulge or a nuclearstar cluster can be fitted with a general S´ersic profile. A profileshowing a increase of (optical) surface brightness in the central part c (cid:13) , 000–000 L.S. Pilyugin et al. of a galaxy (with or without bulge-like component) will be referredto as a steep inner profile below. Irregular galaxies with a steepinner profile are presented in Fig. 2.It is known that the surface brightnesses in some irregulargalaxies are flat or even increase out to a region of slope changewhere they tend to fall o ff (Swaters & Balcells 2002; Taylor et al.2005; Herrmann et al. 2013). Such surface brightness profiles canbe formally fitted by an exponential disk with a bulge-like compo-nent of negative brightness. Such profiles will be referred to as flatinner profiles below. Irregular galaxies with flat inner profiles arepresented in Fig. 3.To define the type of surface brightness profile we use oursurface brightness profiles in the SDSS r band or R -band profilesfrom Swaters & Balcells (2002). It should be noted that the shapesof the surface brightness profiles of the same galaxy in the di ff erentphotometric bands do not necessarily coincide with each other. We determine the T e -based oxygen (O / H) T e and nitrogen (N / H) T e abundances in H ii regions where the auroral line [O iii ] λ T e -method from Pilyugin et al.(2010, 2012).A new method (called the “ C method”) for oxygen and nitro-gen abundance determinations from strong emission lines has re-cently been suggested (Pilyugin et al. 2012, 2013). Here, the stronglines R = [O iii ] λλ N = [N ii ] λλ S = [S ii ] λλ / H) C NS and nitrogen (N / H) C NS abundances in individual H ii re-gions of our target galaxies. The deprojected radii of the H ii regions were computed using theircoordinates and geometrical parameters (position angle of the ma-jor axis and galaxy inclination) listed in Table 1.The radial oxygen abundance distribution within the isophotalradius in every galaxy was fitted by the following equation:12 + log(O / H) = + log(O / H) R + C O / H × ( R / R ) , (1)where 12 + log(O / H) R is the oxygen abundance at R =
0, i.e.,the extrapolated central oxygen abundance. C O / H is the slope ofthe oxygen abundance gradient expressed in terms of dex R − , and R / R is the fractional radius (the galactocentric distance normal-ized to the disk’s isophotal radius R ). The derived parameters ofthe oxygen abundance distributions are presented in Table 2. Thename of the galaxy is listed in column 1. The optical isophotal ra-dius R in kpc is reported in column 2. The extrapolated central12 + log(O / H) R oxygen abundance and the gradient expressed interms of dex R − are listed in columns 3 and 4 (the bootstrappederror of the gradient is given in parenthesis). The scatter of oxygenabundances around the general radial oxygen abundance trend isreported in column 5. The references to sources for spectroscopicdata are given in column 9 . The radial distributions of the oxygenabundances in irregular galaxies are shown in Figs. 2 and 3 togetherwith the surface brightness profiles.The statistical error of the gradient listed in column 4 comesfrom the best fitting procedure. We also estimate the bootstrapped error of the gradient in the following way. The measured H ii re-gions in a galaxy are numbered from 1 to n . We then produce n random integer numbers using a random number generator, andform a bootstrapped subsample of H ii regions choosing the corre-sponding H ii regions from the original sample of H ii regions. Theamount of H ii regions in the bootstrapped subsample is adoptedto be equal to the amount of the H ii regions in the original sam-ple. Thus, some H ii regions from the original sample can be re-peatedly included in the bootstrapped subsample while other H ii regions from the original sample will not at all be included in thebootstrapped subsample. If a bootstrapped subsample involves lessthan three di ff erent H ii regions then this subsample is rejected. Theabundance gradient for the bootstrapped subsample is determinedthrough the best fit, and the error of the original gradient, i.e., thedi ff erence between the values of the gradients for the bootstrappedsubsample and for the original sample of H ii regions is obtained.We considered k = bootstrapped subsamples and determinedthe bootstrapped error of the gradient as ([( P di f f erence j ) / k ] / ).This bootstrapped error of the oxygen abundance gradient is givenin Table 2, column 4 in parenthesis.The statistical and bootstrapped errors of the oxygen abun-dance gradients are close to each other except in the case of thegalaxy UGC 2216 where the bootstrapped error exceeds dramati-cally the statistical error. This is caused by the following. The ra-dial abundance gradient in the UGC 2216 is strongly biased by anH ii region at a galactocentric distance of 4.43 kpc. When the boot-strapped subsample does not contain this point then the value ofthe radial abundance gradient is very uncertain since in this casethe gradient is determined from measurements at close galactocen-tric distances. As a result, the bootstrapped error of the radial abun-dance gradient for this galaxy is quite large.As in the case of the oxygen abundance, the radial nitrogenabundance distribution in every galaxy was fitted by the followingequation:12 + log(N / H) = + log(N / H) R + C N / H × ( R / R ) . (2)The derived parameters of the nitrogen abundance distributions arepresented in Table 2. The extrapolated central 12 + log(N / H) R ni-trogen abundance and the gradient in terms of dex R − are listedin columns 6 and 7. The scatter of oxygen abundances around thegeneral radial oxygen abundance trend is reported in column 8. Thevalue in the parenthesis in column 7 is the bootstrapped error of thenitrogen abundance gradient obtained in the same way as for theoxygen abundance gradient.The radial oxygen abundance gradients in irregular galaxiesobtained here are based mainly (or only) on oxygen abundances(O / H) C NS estimated through strong emission lines using the C NS method. Figs. 2 and 3 show that the scatter in the (O / H) C NS abun-dances around the general radial trend is often lower than the scatterin the (O / H) T e abundances. Five galaxies from our present sam-ple are in the list of galaxies considered in our previous study(Pilyugin et al. 2014a). The values of gradients obtained here areslightly di ff erent from those reported in our previous study for thefollowing reasons. First, in our current work we obtain and use newparameters for our target galaxies such as inclination, position an-gle of the major axis, and optical isophotal radius. Furthermore, inPilyugin et al. (2014a) the oxygen and nitrogen abundances wereestimated via the C ON method for H ii regions with available mea-surements of the [O ii ] λλ C NS method for the other H ii regions. In our current study, the oxygenand nitrogen abundances were estimated through the C NS methodfor all H ii regions, and T e -based abundances are added. c (cid:13) , 000–000 bundance gradients in irregulars log Σ L NGC 4214 r (kpc) NGC 4214 log Σ L NGC 4395 r (kpc) NGC 4395 log Σ L UGC 7557 r (kpc) UGC 7557 log Σ L NGC 4449 r (kpc) NGC 4449 log Σ L CG 071-090 r (kpc) CG 071-090 log Σ L UGC 9614 r (kpc) UGC 09614 log Σ L UGC 12709 r (kpc) UGC 12709 Figure 2.
Surface brightness profiles and radial distributions of oxygen abundances for irregular galaxies with steep inner profiles. Each galaxy is presentedin two panels. Each upper panel N a shows the surface brightness profiles in the SDSS g band as a light-grey (blue) solid line, in the SDSS r band as a dark(black) long-dashed line, and in the WISE W N b shows the oxygen abundance in individualH ii regions as a function of radius. The dark (black) open circles show (O / H) C NS abundances and the grey (red) filled circles indicate the (O / H) T e abundances.The solid line represents the inferred linear abundance gradient. (A color version of this figure is available in the on-line edition.) Fig. 2 shows the surface brightness profiles and radial distributionsof the oxygen abundances for the irregular galaxies with steep innerprofiles. Each galaxy is presented in two panels. Each upper panel N a shows the surface brightness profiles in the SDSS g band as alight-grey (blue) solid line, in the SDSS r band as a dark (black)long-dashed line, and in the WISE W N b shows the oxygen abun-dance in individual H ii regions (open circles) as a function of ra-dius. The linear best fit to those data is indicated by a solid line.Fig. 2 shows that irregular galaxies with steep inner profiles haveappreciable radial abundance gradients.Fig. 3 shows the surface brightness profiles and radial distribu-tions of the oxygen abundances for irregular galaxies with flat innerprofiles. Inspection of Fig. 3 shows that the radial abundance gra- c (cid:13) , 000–000 L.S. Pilyugin et al. log Σ L UGC 02023 r (kpc) UGC 02023 log Σ L UGC 02216 r (kpc) UGC 02216 log Σ L NGC 1156 r (kpc) NGC 1156 log Σ L NGC 2537 r (kpc) NGC 2537 log Σ L UGC 04305 r (kpc) UGC 04305 log Σ L NGC 3738 r (kpc) NGC 3738 log Σ L UGC 6980 r (kpc) UGC 6980 Figure 3.
The same as Fig. 2 but for irregular galaxies with flat inner profiles. For galaxies without SDSS photometric maps, the surface brightness profile inthe R band from Swaters & Balcells (2002) is indicated with dark (black) plus signs. (A color version of this figure is available in the on-line edition.) dients in the irregular galaxies with flat inner profiles are shallowerthan the gradients in irregular galaxies with steep inner profiles.Thus, our data suggest that there is a relation between theradial abundance gradient in an irregular galaxy and its surfacebrightness profile. Panel a of Fig. 4 shows the radial oxygen abun-dance gradient as a function of optical radius R for our sampleof irregular galaxies. The dark (black) open circles mark irregu-lar galaxies with steep inner photometric profiles. The dark-grey(red) open squares denote galaxies with flat inner profiles. The dark(black) dotted line is the arithmetic mean of the gradients for galax- ies with steep inner photometric profiles, whereas the dark-grey(red) dashed line is the mean for galaxies with flat inner photomet-ric profiles. The light-grey (green) solid line is the arithmetic meanof the gradients for all our galaxies (both those with steep and thosewith flat inner profiles). Since the numbers of galaxies in our sam-ples are small even one deviant galaxy may appreciably changethe arithmetic mean for the sample. Indeed the aritmetic mean ofthe gradients for galaxies with steep inner photometric profiles ischanged by ∼ R − when the deviating galaxy NGC 4214 c (cid:13) , 000–000 bundance gradients in irregulars Table 2.
The derived parameters of the radial oxygen and nitrogen abundance distributions in our target galaxies.Galaxy R + log(O / H) R O / H gradient σ (O / H) 12 + log(N / H) R N / H gradient σ (N / H) Referenceskpc dex R − dex dex R − dexUGC 02023 2.24 8.08 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ff man (1999), 11 – Kobulnicky & Skillman (1996), 12 –Lequeux et al. (1979), 13 – McCall et al. (1985), 14 – Moustakas & Kennicutt (2006), 15 – Romanishin et al. (1983), 16 – SDSS (York et al. (2000), 17 –van Zee et al. (1998), 18 – van Zee & Haynes (2006). (with a positive gradient 0.049 dex R − ) is excluded from consider-ation.Panel b of Fig. 4 shows the same as panel a but for the ni-trogen abundance gradients. Comparison between panels a and b shows that the general picture is similar for oxygen and nitrogenabundance gradients, i.e., the irregular galaxies with flat inner pho-tometric profiles have shallower nitrogen abundance gradients ascompared to galaxies with steep inner photometric profiles.Panel c of Fig. 4 shows the radial oxygen abundance gradientswith bootstrapped errors (reported in Table 2 in parenthesis). Thefilled dark (black) circle on the right side of the panel shows themean value of the gradients within its 95% and 68% confidence in-tervals for the sample of galaxies with steep inner photometric pro-files. To estimate the confidence interval of the mean value of thegradients of the sample of galaxies the bootstrap method is used.We create 10 bootstrapped subsamples from the original sampleof gradients keeping the size of each bootstrapped subsample equalto the size of the original sample, and modifying the value of theoriginal gradient of each galaxy by introducing a random error. Thiserror is randomly chosen from a set of errors that follow a Gaus-sian distribution scaled to the standard deviation corresponding tothe bootstrapped error of abundance gradient (reported in Table 2in parenthesis). We consider the distribution of the mean valuesof the abundance gradients for those 10 bootstrapped subsamplesand determine the 95% and 68% confidence intervals of the meanabundance gradient for the sample of galaxies. The filled dark-grey(red) square shows such a mean value of the abundance gradientsfor the sample of galaxies with flat inner photometric profiles, andthe light-grey (green) asterisk shows the one for the total sample ofgalaxies. Panel d of Fig. 4 shows the same as panel c but for theradial nitrogen abundance gradients.The di ff erence between the mean values of the oxygen abun-dance gradients for galaxies with steep and flat inner photometricprofiles is estimated in a similar way and amounts to − .
126 dex R − within the 95% confidence interval ( − . − .
236 dex R − within the 95% confidence interval ( − . ff erence between the mean values of the abundancegradients in irregular galaxies with steep and flat inner photometricprofiles exists (is less than 0) at 91% confidence level for oxygenabundance gradients and at 94% confidence level for nitrogen abun-dance gradients.Thus, our data suggest that i) there are radial abundance gradi-ents in irregular galaxies, and ii) there is a di ff erence between radialabundance gradients in irregular galaxies with steep and flat innerphotometric profiles with a probability higher than 90%.It should be noted that here the abundances are determinedthrough the C and T e methods. The C method is based on the abun-dances derived via the T e method and, consequently, produces theabundances on the same metallicity scale as the T e method. If theabundances derived using the T e method are not correct for somereason (e.g., because of small-scale temperature fluctuations withinan H ii region (Peimbert 1967), or if the energies of the electrons inan H ii region do not follow a Maxwell distribution (Dopita et al.2013)) then our abundances should be revised. Furthermore, theabsolute metallicity scale of H ii regions varies up to ∼ ii regions but only within a limited interval. The oxygenabundances of irregular galaxies typically are within or near thetransition zone in the R – O / H diagram where previous calibra-tions cannot be used or where they provide abundances with largeuncertainties. Therefore the T e - and C -based abundances are prefer-able for irregular galaxies.It is known (e.g., Searle & Sargent 1972; Pagel 1997) that theradial distribution of oxygen abundances in the disk of a galaxy iscontrolled by the variation of the astration level (or gas mass frac-tion µ ) with radius and by the mass exchange between a galaxyand the surrounding medium (via galactic winds and / or gas in-fall / merging) and between di ff erent parts of a galaxy. Taking intoconsideration the radial variation of the astration level, one mayexpect that physical gradients (expressed in dex kpc − ) in irregu-lar galaxies can be even steeper than those in spiral galaxies. The c (cid:13) , 000–000 L.S. Pilyugin et al. -1.0-0.8-0.6-0.4-0.20.00.20.40.6 1 2 3 4 5 6 7 8 9 grad(O/H) dex/R R kpc a steep inner profileflat ihher profile -1.0-0.8-0.6-0.4-0.20.00.20.40.6 1 2 3 4 5 6 7 8 9 grad(N/H) dex/R R kpc b steep inner profileflat ihher profile -1.0-0.8-0.6-0.4-0.20.00.20.40.6 0 1 2 3 4 5 6 7 8 9 grad(O/H) dex/R R kpc c -1.0-0.8-0.6-0.4-0.20.00.20.40.6 0 1 2 3 4 5 6 7 8 9 grad(N/H) dex/R R kpc d Figure 4.
The panel a shows the oxygen abundance gradient as a function of optical radius R . The dark (black) open circles mark irregular galaxies withsteep inner photometric profiles. The dark-grey (red) open squares denote galaxies with flat inner photometric profiles. The dark (black) dotted line is thearithmetic mean of the gradients for galaxies with steep inner profiles, the dark-grey (red) dashed line for galaxies with flat inner profiles, and the light-grey(green) solid line the total sample. Panel b shows the same as panel a but for the nitrogen abundance gradients. Panel c shows the oxygen abundance gradientswith bootstrapped errors. On the right side of the panel, the mean values of the gradients for the sample of galaxies with steep inner photometric profiles (thefilled dark (black) circle), for the sample of galaxies with flat inner photometric profiles (the filled dark-grey (red) square), and for total sample (the ligh-grey(green) asterisk) and their 95% and 68% confidence intervals are shown. Panel d shows the same as panel c but for the nitrogen abundance gradients. (A colorversion of this figure is available in the on-line edition.) metallicities in irregular galaxies are typically lower than the onesin spiral galaxies since irregular galaxies are less massive and lessevolved. The simple model for the chemical evolution of galaxiespredicts that the oxygen abundance O / H varies with gas mass frac-tion µ more strongly at low metallicity. Thus a similar change of µ along the radial direction would result in a larger change of O / H inirregular galaxies than in spiral galaxies.Radial mixing of gas flattens the abundance gradient in thedisk of a galaxy. Radial mixing of gas can be caused by interact-ing or merging galaxies (e.g., Rupke et al. 2010a,b) and by galac-tic fountains (galactic winds and subsequent gas infall). The argu-ments pro and contra galactic wind-dominated evolution of irregu-lar galaxies are discussed in many studies devoted to the chemical evolution of galaxies (Skillman 1997; Cavil´an et al. 2013, amongmany others). A galactic wind can be caused by the injection ofenergy by multiple, spatially and temporally clustered supernovaein a galaxy undergoing a starburst (De Young & Gallagher 1990;Mac Low & Ferrara 1999). The e ffi ciency of the galactic winds de-pends on the number of massive stars that are progenitors of super-novae in a star formation event. Lee et al. (2009) found that con-tinuous, steady star formation dominates in the present epoch indwarf galaxies. Only ∼
6% of low-mass galaxies experience strongstar formation bursts. The fraction of stars formed in starbursts is ∼ c (cid:13) , 000–000 bundance gradients in irregulars Thus, we can interpret our results in the following manner. Ir-regular galaxies with steep inner profiles do not seem to undergostrong radial mixing of gas at the present epoch and show consider-able radial abundance gradients. The radial mixing of gas (throughradial flows or galactic fountains) took place in irregular galaxieswith flat inner profiles, resulting in shallower (if any) gradients ascompared to the galaxies with steep inner profiles. It should benoted that the physical reason for di ff erent radial profile types isstill a mystery. It is not even clear why there is an exponential drop-o ff of the brightness profile (Herrmann et al. 2013). We determined the abundance distributions traced by H ii regionsand compare their shape with the surface brightness profiles of thedisks of fourteen irregular Sm and Im galaxies (morphological T types of T = T = ii regions from di ff erent studies to infer theabundances. The oxygen (O / H) T e and nitrogen (N / H) T e abundancesin the H ii regions with the detected auroral line [O iii ] λ T e -method. In theother H ii regions, oxygen (O / H) C NS and nitrogen (N / H) C NS abun-dances were obtained through the C method. We then quantifiedthe values of the gradients of the radial abundance profiles.Moreover, we constructed radial surface brightness profiles inthe infrared W g and r bands usingthe publicly available photometric maps. The irregular galaxies ofour sample can be divided into two types according to the shapesof their surface brightness profiles: those with steep inner profiles,and those with flat inner profiles.We find that there is a correspondence between the radialabundance gradient in an irregular galaxy and its surface bright-ness profile with a probability higher than 90%. Irregular galaxieswith steep inner profiles usually show a considerable radial abun-dance gradient. Irregular galaxies with flat inner surface brightnessprofiles have shallower gradients (if any) as compared to galaxieswith steep inner profiles.Thus, irregular galaxies with steep inner profiles show usu-ally a pronounced radial abundance gradient that resembles that ofspiral galaxies. In that sense, those irregular galaxies seem to ex-tend the Hubble sequence of spiral galaxies. In other words, ourdata suggest that there is no “spiral versus irregular dichotomy” interms of radial abundance gradients existing only in spiral galaxies,but not in irregulars. While irregulars have long been believed to bechemically homogeneous, our study shows that given enough mea-surements of nebular abundances of H ii regions across a wide rangeof galactocentric radii, irregulars may well exhibit radial abundancegradients. This tendency is particularly conspicuous in irregularswith steep surface brightness profiles in their inner regions. ACKNOWLEDGEMENTS
We are grateful to the referee for his / her constructive comments.L.S.P., E.K.G., and I.A.Z. acknowledge support within the frame-work of Sonderforschungsbereich (SFB 881) on ”The Milky WaySystem” (especially subproject A5), which is funded by the Ger-man Research Foundation (DFG).L.S.P. and I.A.Z thank the hospitality of the AstronomischesRechen-Institut at Heidelberg University where part of this investi-gation was carried out. This work was partly funded by the subsidy allocated to Kazan Fed-eral University for the state assignment in the sphere of scientificactivities (L.S.P.).We thank R.A. Swaters and M. Balcells for supporting us with thesurface brightnes profiles of galaxies from their sample in numeri-cal form.This research made use of Montage, funded by the National Aero-nautics and Space Administration’s Earth Science TechnologyO ffi ce, Computational Technnologies Project, under CooperativeAgreement Number NCC5-626 between NASA and the CaliforniaInstitute of Technology. The code is maintained by the NASA / IPACInfrared Science Archive.Funding for the SDSS and SDSS-II has been provided by the Al-fred P. Sloan Foundation, the Participating Institutions, the Na-tional Science Foundation, the U.S. Department of Energy, theNational Aeronautics and Space Administration, the JapaneseMonbukagakusho, the Max Planck Society, and the Higher Ed-ucation Funding Council for England. The SDSS Web Site ishttp: // / . REFERENCES
Ahn C.P. et al., 2012, ApJS, 203, 21Alloin D., Collin-Sou ff rin S., Joly M., Vigroux L., 1979, A&A,78, 200Berg D.A. et al., 2012, ApJ, 754, 98Blanton M.R., Roweis S., 2007, AJ, 133, 734B¨oker T., van der Marel R.P., Mazzuca L., Rix H.-W., Rudnick G.,Ho L.C., Shields J.G., 2001, AJ, 121, 1473Cardelli J.A., Clayton G.C., Mathis J.S., 1989, ApJ, 345, 245Casagrande L., Ram´ırez I., Mel´endez J., Asplund M., 2012, ApJ,761, 16Cavil´an M., Ascasibar Y., Moll´a M., D´ıaz ´A.I., 2013, MNRAS,434, 2491Cignoni M., Cole A.A., Tosi M., Gallagher JS., Sabbi E., Ander-son J., Grebel E.K., Nota A., 2013, ApJ, 775, 83Croxall K.V., van Zee l., Lee H., Skillman E.D., Lee J., Cˆot´e S.,Kennicutt R.C., Miller B.W., 2009, ApJ, 705, 723de Vaucouleurs G., de Vaucouleurs A., Corvin H.G., Buta R.J.,Paturel G., Fouque P., 1991, Third Reference Catalog of BrightGalaxies, New York: Springer Verlag (RC3)De Young D.S., Gallagher J.S., 1990, ApJ, 356, L15Dopita M.A., Sutherland R.S., Nicolls D.C., Kewley L.S., VoghtF.P.A., 2013, ApJS, 208, 10Egorov O.A., Lozinskaya T.A., Moiseev A.V., 2013, MNRAS,429, 1450Esteban C., Bresolin F., Peimbert M., Garc´ıa-Rojas J., PeimbertA., Mesa-Delgado A., 2009, ApJ, 700, 654Garnett D.R., 2002, ApJ, 581, 1019Gil de Paz A., Zamorano J., Gallego J., Dom´ıngues F. de B., 2000,A&AS, 145, 377Glatt K. et al., 2008, AJ, 136, 1703Gusev A.S., Pilyugin L.S., Sakhibov F., Dodonov S.N., EzhkovaO.V., Khramtsova M.S., 2012, MNRAS, 424, 1930Haurberg N.C., Rosenberg J., Salzer J.J., 2013, ApJ, 765, 66Herrmann K.A., Hunter D.A., Elmegreen B.G., 2013, AJ, 146,104Ho L.C., Filippenko A.V., Sargent W.L.W., 1997, ApJS, 112, 315Hunter D.A., Ho ff man L., 1999, AJ, 117, 2789Hunter D.A., Gallagher J.S., Rautenkranz D., 1982, ApJS, 49, 53Karachentsev I.D., Makarov D.I., Kaisina E.I., 2013, AJ, 145, 10 c (cid:13) , 000–000 L.S. Pilyugin et al.
Kewley L.J., Ellison S.L., 2008, ApJ, 681, 1183Kniazev A.Y., Grebel E.K., Pustilnik S.A., Pramskij A.G., ZuckerD.B., 2005, AJ, 130, 1558Kobulnicky H.A., Skillman E.D., 1996, ApJ, 471, 211Lee H., Zucker D.B., Grebel E.K., 2007, MNRAS, 376, 820Lee J.C., Kennicutt R.C., Funes J.G., Sakai S., Akiyama S., 2009,ApJ, 692, 1305Lequeux J., Peimbert M., Rayo J.F., Serrano A., Torres-PeimbertS., 1979, A&A, 80, 155Mac Low M.-M., Ferrara A., 1999, ApJ, 219, 46McCall M.L., Rybski P.M., Shields G.A., 1985, ApJS, 57, 1Miller B.M., Hodge P., 1996, ApJ, 458, 467Moustakas J., Kennicutt R.C., 2006, ApJS, 164, 81Moustakas J., Kennicutt R.C., Tremonti C.A., Dale D.A., SmithJ.-D.T., Calzetti D., 2010, ApJS, 190, 233Pagel B.E.J. 1997, Nucleosynthesis and Chemical Evolution ofGalaxies (Cambridge: Cambridge Univ. Press)Pagel B.E.J., Edmunds M.G., Fosbury R.A.E., Webster B.L.,1978, MNRAS, 184, 569Pagel B.E.J., Edmunds M.G., Blackwell D.E., Chun M.S., SmithG., 1979, MNRAS, 189, 95Peimbert M., 1967, ApJ, 150, 825Pilyugin L.S., Ferrini F., 2000, A&A, 358, 72Pilyugin L.S., V´ılchez J.M., Contini T., 2004, A&A, 425, 849Pilyugin L.S., Thuan T.X., V´ılchez J.M., 2007, MNRAS, 376, 353Pilyugin L.S., V´ılchez J.M., Thuan T.X., 2010, ApJ, 720, 1738Pilyugin L.S., Grebel E.K., Mattsson L., 2012, MNRAS, 424,2316Pilyugin L.S., Lara-L´opez M.A., Grebel E.K., Kehrig C.,Zinchenko I.A., L´opez-S´anchez ´A.R., V´ılchez J.M., Mattsson L.,2013, MNRAS, 432, 1217Pilyugin L.S., Grebel E.K., Kniazev A.Y., 2014a, AJ, 147, 131Pilyugin L.S., Grebel E.K., Zinchenko I.A., Kniazev A.Y., 2014b,AJ, 148, 134Romanishin W., Strom K.M., Strom S.E., 1983, ApJS, 53, 105Roy J.-R., Belley J., Dutil Y., Martin P., 1996, ApJ, 460, 284Rupke D.S.N., Kewley L.J., Barnes J.E., 2010, ApJ, 710, L156Rupke D.S.N., Kewley L.J., Chien L.-H., 2010, ApJ, 723, 1255S´anchez S.F. et al., 2014, A&A, 563, 49Schlafly E.F., Finkbeiner D.P., 2011, ApJ, 737, 103Schlegel D.J., Finkbeiner D.P., Davis M., 1998, ApJ, 500, 525Searle L., Sargent W.L.W., 1972, ApJ, 173, 25Skillman E.D., 1997, RevMexAA (Serie de Conferencias), 6, 36Swaters R.A., Balcells M., 2002, A&A, 390, 863Taylor V.A., Jansen R.A., Windhorst R.A., Odewahn S.C., Hib-bard J.E., 2005, ApJ, 630, 784van Zee L., Haynes M.P., 2006, ApJ, 636, 214van Zee L., Salzer J.J., Haynes M.P., O‘Donoghue A. A., BalonekT.J., 1998, AJ, 116, 2805Vila-Costas M.B., Edmunds M.G. 1992, MNRAS, 259, 121Wright E.L. et al., 2010, AJ, 140, 1868York D.G. et al., 2000, AJ, 120, 1579Zaritsky D., Kennicutt R.C., Huchra, J.P., 1994, ApJ, 420, 87 c (cid:13)000