Clues on the rejuvenation of the S0 galaxy NGC 404 from the chemical abundance of its outer disk
PPreprint typeset using L A TEX style emulateapj v. 5/2/11
CLUES ON THE REJUVENATION OF THE S0 GALAXY NGC 404 FROM THE CHEMICAL ABUNDANCE OF ITSOUTER DISK F ABIO B RESOLIN
Institute for Astronomy, 2680 Woodlawn Drive, Honolulu, HI [email protected]
ABSTRACTThe oxygen abundance of the outer disk of the nearby S0 galaxy NGC 404, a prototypical early-type galaxywith extended star formation, has been derived from the analysis of H II region spectra. The high mean valuefound, 12 + log(O/H) = 8 . ± .
1, equivalent to approximately 80% of the solar value, argues against both thepreviously proposed cold accretion and recent merger scenarios as viable mechanisms for the assembly of thestar-forming gas. The combination of the present-day gas metallicity with the published star formation historyof this galaxy favors a model in which the recent star forming activity represents the declining tail of the originalone.
Subject headings: galaxies: abundances — galaxies: ISM — galaxies: elliptical and lenticular, cD — galaxies:evolution — galaxies: individual (NGC 404) INTRODUCTION
Studies of the gaseous content of early-type (E/S0) galaxies(ETGs) have established that gas accretion and the presenceof H I reservoirs are common features among these systems(Oosterloo et al. 2010; Davis et al. 2011; Thom et al. 2012).The presence of gas is often accompanied by low-level, recent(age < > z ∼ . I ring, ex-tending between 1.5 and 6.5 kpc (1 – 4 R ) from the galacticcenter. The relatively small distance to NGC 404 and its isola-tion (Karachentsev et al. 2002) make this galaxy an ideal sitewhere to explore the processes regulating the evolution of anETG with extended star formation. Following the suggestionby del Río et al. (2004) that a merger with a dwarf galaxy sup-plied NGC 404 with fresh gas in the past Gyr, Thilker et al.(2010) argued that this lenticular galaxy is an example of arejuvenated ETG, currently shifting from the red sequence ofthe bimodal optical color distribution of galaxies back into thegreen valley. This paper offers an alternative view, provid-ing constraints on the nature of the extended star formation inNGC 404 from a chemical abundance analysis of H II regionslying in its outer disk. OBSERVATIONS AND DATA REDUCTION
Observations of H II region candidates in the outer disk ofNGC 404 were obtained with the Gemini Multi-Object Spec-trograph (GMOS, Hook et al. 2004) at the Gemini North tele-scope. The targets were selected from H α narrow-band im-ages of two 5 . (cid:48) × . (cid:48) . (cid:48) . (cid:48) . (cid:48)(cid:48) . (cid:48)(cid:48) ∼ IRAF routines in the gemini/gmos package were used forelectronic bias subtraction, flat field correction, wavelengthcalibration, and coaddition of the raw data frames. The spec- IRAF is distributed by the National Optical Astronomy Observatories,which are operated by the Association of Universities for Research in As-tronomy, Inc., under cooperative agreement with the National Science Foun-dation. a r X i v : . [ a s t r o - ph . C O ] J un Bresolin
TABLE 1H II region sample: coordinates and line ratiosID R.A. DEC R R / R [O III ] λ β [N II ] λ α [S II ] λ α Comments(J2000.0) (J2000.0) (arcsec)1 01 09 18.65 35 41 57.5 124 1.18 < < ··· < < < ··· ··· ··· ··· < <
12 34 567 8910/11 121314 15 161718 19 2 arcmin F IG . 1.— Identification of the targets on a near-UV GALEX image ofNGC 404 (N at top, E on the left). The circles, drawn at radii R and 4 R ,represent the approximate inner and outer boundaries of the star forming ring. tra were not flux calibrated. This has virtually no effect on theanalysis presented here, given the proximity in wavelengthbetween the emission lines used to construct the nebular di-agnostics. Two of the targets were found to be backgroundemission-line galaxies, and were removed from the analysis.The final sample comprises 19 objects in an annulus between R and 4 R (hereafter referred to as the star-forming ring) ofNGC 404. Their location is indicated in Fig. 1 ( R = 1 . (cid:48) CHEMICAL ABUNDANCE ANALYSIS
The emission lines covered in the observed spectral rangeinclude H α , [N II ] λ II ] λλ β and [O III ] λ III ] λ β ,[N II ] λ α and [S II ] λλ α , which are virtu-ally independent of flux calibration and reddening corrections,given the small separation in wavelength of the lines involved.Fig. 2 illustrates the excitation properties of the targets,showing their location in the [N II ]/H α vs. [O III ]/H β and[S II ]/H α vs. [N II ]/H α diagrams. Three objects stand out aspeculiar based on their line ratios: II regions, whosechemical analysis can be carried out using standard diagnos-tics.The diagrams in Fig. 2 include curves representing H II re-gion models from Dopita et al. (2006), calculated for ionizingcluster ages between 0.1 and 4 Myr, and different metallici-ties (0.2, 0.4, 1.0 and 2.0 × Z (cid:12) ). The parameter R ∝ M cl / P (the ratio between cluster mass and pressure of the interstel-lar medium) is also varied between log R = − R = 0 (dashed lines), although this has only a sec-ondary effect on the interpretation of the diagrams. Accord-ing to these models, the bulk of the H II region sample hasa metallicity slightly below solar. Since the model curves inthe [N II ]/H α vs. [O III ]/H β diagram become virtually verti-cal after a cluster age of 3 Myr, the same conclusion can alsobe drawn for the 5 objects having only an upper limit for the[O III ]/H β ratio.In order to quantify the radial distribution of thenebular metallicities the abundance diagnostic N2 =log([N II ] λ α ) was adopted, using the calibrations pro-vided by Pettini & Pagel (2004, = PP04) and Denicoló et al.(2002, = D02). The resulting O/H abundances are presented inFig. 3, using solid (PP04) and open (D02) circles. While mostsources share a similar O/H ratio, F IG . 2.— Excitation diagrams: [N II ]/H α vs. [O III ]/H β (left) and [S II ]/H α vs. [N II ]/H α (right) . Open circles: measured flux ratios. The candidate planetarynebulae ( . − × Z (cid:12) , log R = − R = 0 (dotted lines). largest H α equivalent width (950 Å) in the sample, indicat-ing a very young nebula, which is supported by the models inFig. 2 (point lying on the 0.1 Myr curve). The position of II ]/H α ratio (which leads to the anomalous low O/H ratiousing the strong-line abundance diagnostics), its metallicity isin fact comparable to that of the rest of the sample.A least-squares fit to the data points (excluding theoutlier .
55 ( ± . − .
014 ( ± . R / R . The (virtually flat) gradient’s slopeis unchanged using the D02 calibration, and can be expressedas d(O/H)/dR = − .
009 ( ± . − . The mean abun-dance values are (cid:104)
12 + log(O/H) (cid:105) PP = 8 .
52 ( ± .
03) and (cid:104)
12 + log(O/H) (cid:105) D = 8 .
63 ( ± . III ] λ β ) − N2} diagnostic (PP04)based on [O
III ] λ direct method, tends to yield lower O/H ratios comparedto alternative strong-line abundance determination methods(Bresolin et al. 2009a). In summary, the diagnostics con-sidered here indicate that the oxygen abundance in the outerdisk of NGC 404 is essentially constant, and I adopt a repre-sentative mean value 12 + log(O/H) = 8.6 ± .
1, equivalent to ∼ × the solar value (Asplund et al. 2009). This result isconsistent with the information provided by the photoioniza-tion models shown in Fig. 2. F IG . 3.— Radial distribution of the O/H abundance ratio obtained from theN2 diagnostic, using the calibrations of Pettini & Pagel (2004, = PP04) andDenicoló et al. (2002, = D02). The full line represents a least-squares fit to thedata points measured with the PP04 calibration. The dashed lines show themean values (indicated) of the O/H ratio resulting from the two calibrations.Triangles and arrows correspond to the abundances measured from the O3N2diagnostic based on [O III ] λ DISCUSSION
The star formation history of NGC 404 has been recently in-vestigated by Williams et al. (2010) from theoretical fits to thecolor-magnitude diagrams (CMDs) of the stellar populationsin three WFPC2 fields observed with the
Hubble Space Tele-scope . These fields cover the galactocentric distance range0.61 – 2.55 R , overlapping with the range spanned by theH II regions observed in this work (1.06 – 3.66 R ). TheCMD analysis indicates that the majority (70%) of the starsin the disk (even in the outer star forming ring) formed ear-lier than 10 Gyr ago (90% formed by the end of the following Bresolin2 Gyr). The star formation rate (SFR) surface density declinedfrom ∼ . × − M (cid:12) yr − kpc − early in the history of thegalaxy to a negligibly small value ( (cid:46) − M (cid:12) yr − kpc − )approximately 0.9 Gyr ago, when the gas surface density wasextremely small ( ∼ . M (cid:12) pc − ). The CMD fitting suggeststhat in the recent past ( ∼
500 Myr ago) the star formationactivity resumed, with an estimated SFR density of 2 × − M (cid:12) yr − kpc − in the time interval between 0.5 and 0.2 Gyrago (Williams et al. 2010, Fig. 14). This is about one orderof magnitude higher than the current value estimated from thefar-UV emission of the star-forming ring (Thilker et al. 2010).Williams et al. (2010) speculated that the increase in star for-mation activity was possibly triggered by the merger postu-lated by del Río et al. (2004), who studied with 21 cm obser-vations the morphology and kinematics of the 1 . × M (cid:12) H I outer disk. del Río et al. (2004) identified the source ofthis gas with a merging dwarf galaxy about 0.5-1 Gyr ago,lending credibility to the rejuvenation scenario to explain therecent star formation in NGC 404.Is this picture consistent with the chemical abundance anal-ysis carried out in Sect. 3? It is worth comparing first thepresent-day gas metallicity with the stellar metallicity in thedisk, as inferred by Williams et al. (2010) from the CMDmodeling. As mentioned earlier, most of the galaxy’s gasconverted into stars very early on: the typical stellar age inNGC 404 is 12 Gyr, with a fitted metallicity [M/H] = − . ± .
37. The declining star formation activity caused a progres-sive metal enrichment of the galaxy. The errors in [M/H]given by Williams et al. (2010) become too large ( > <
200 Myr ago) to offer a meaningful com-parison with the H II region chemical abundances, but theseauthors estimated a stellar metallicity [M/H] = − . ± . (cid:39) − . II regions, [O/H] (cid:39) − . ± . α /Fe] (cid:39) II regions provide a determination of the O/H ratio, while thestellar M/H given above is assumed to be representative ofthe Fe/H ratio).If the general description of the star formation history ofNGC 404 outlined by Williams et al. (2010) is broadly cor-rect, the significant chemical enrichment measured for theouter disk H II regions argues against a gas-rich dwarf galaxyas the main source of fuel responsible for the re-ignition ofthe star formation, as proposed by del Río et al. (2004) andThilker et al. (2010). From the stellar mass of NGC 404 mea-sured by Thilker et al. (2010), and assuming a mass mergerratio of 1:3 (the maximum ratio commonly adopted to distin-guish between major and minor mergers), the stellar mass ofthe merging dwarf would be 1 . × M (cid:12) . According to themass-metallicity relation observed in the local universe (Berget al. 2012) a galaxy with this mass has an oxygen abundance12 + log(O/H) (cid:39) . ± . Allowing for star formation over a period of 0.3 Gyr ata rate of 2 × − M (cid:12) yr − kpc − (following Williams et al.2010), considering an observed H I surface density in the star-forming ring Σ H I = 1 M (cid:12) pc − (del Río et al. 2004) and with The gas metallicities obtained from the N2 method used here are on thesame absolute scale as the direct metallicities measured by Berg et al. (2012). an initial oxygen abundance 12 + log(O/H) (cid:39) (cid:39) (cid:39) I structure, which suggests that the gas has beenacquired recently, and on critical gas density arguments (in-validated by the GALEX detection of recent star formation).It is conceivable that a perturbing event (minor/dry merger)took place in the recent past, that revitalized the star forma-tion activity, but it is unlikely, based on the reconstructed starformation history, that such an event was responsible for theaccretion of the gas, because of the incompatible chemicalcomposition. It is also possible to argue in favor of an en-riched accretion mechanism, as proposed in recent studies ofthe chemical abundances of the outer disks of spiral galaxies(Bresolin et al. 2009b, 2012; but I also note that these struc-tures are 3 × larger than the outer disk of NGC 404), wherethe gas, chemically enriched by stellar processes in the disk,is subject to a wind-recycling mechanism, and inflows backinto the disk (Davé et al. 2011; Martin et al. 2012).In conclusion, this work shows that the chemical abundanceanalysis of the ionized gas provides crucial constraints on theorigin of the outer star forming rings and potential rejuve-nation events observed in ETGs. In particular, two of themain scenarios proposed, cold accretion and minor merger,are ruled out in the case of NGC 404 by the high nebularmetallicity observed. Future work will test whether the de-clining star formation scenario provides a better fit to thepresent-day gas metallicity of additional ETGs.FB gratefully acknowledges the support from the Na-tional Science Foundation grant AST-1008798, and thanks theanonymous referee for constructive comments.he rejuvenation of the S0 galaxy NGC 404 5larger than the outer disk of NGC 404), wherethe gas, chemically enriched by stellar processes in the disk,is subject to a wind-recycling mechanism, and inflows backinto the disk (Davé et al. 2011; Martin et al. 2012).In conclusion, this work shows that the chemical abundanceanalysis of the ionized gas provides crucial constraints on theorigin of the outer star forming rings and potential rejuve-nation events observed in ETGs. In particular, two of themain scenarios proposed, cold accretion and minor merger,are ruled out in the case of NGC 404 by the high nebularmetallicity observed. Future work will test whether the de-clining star formation scenario provides a better fit to thepresent-day gas metallicity of additional ETGs.FB gratefully acknowledges the support from the Na-tional Science Foundation grant AST-1008798, and thanks theanonymous referee for constructive comments.he rejuvenation of the S0 galaxy NGC 404 5