Nearby early-type galaxies with ionized gas. Characterization of the underlying stellar population

0.13 MB·English

by  F. Annibali

Nearby early–type galaxies with ionized gas.
Characterization of the underlying stellar population
5
0
P
0
2 FrancescaAnnibali∗, LuigiDanese o

n
SISSA,Italy S
a
J E-mail: [email protected];[email protected]
(

4
1 RobertoRampazzo,AlessandroBressan B

1 OsservatorioAstronomicodiPadova,Italy D
E-mail: [email protected];[email protected]
v
M
2
0 EmanueleBertone,MiguelChavez
H
3 INAOE,Puebla,Mexico
1 2
E-mail: [email protected];[email protected]
0
5 0
WernerW.Zeilinger
0
/ InstitutfürAstronomiederUniversitätWien,Austria 0
h
p E-mail: [email protected] 4
-
o )
r WepresentapreliminaryanalysisofthesampleofearlytypegalaxiesofRampazzoetal. (2005),
t 0
s selected to build a data-set of spectral properties of well studied early-type galaxies showing
a
: emission lines. Because of the presence of emission lines, the sample is biased toward objects 5
v
i thatmightbe expectedto have ongoingandrecentstar formation. We havecomparedthe line- 6
X
strengthindicespresentedinRampazzoetal. (2005)withSimpleStellarPopulations(SSPs) in
r
a
order to characterize the underlying stellar population of the galaxies. We have derived ages,
metallicitiesand[a /Fe]ratios. Thepositivetrendofthes -metallicityands −[a /Fe]relations
isreproduced. Thebulkofthegalaxiesspanarangeinmetallicityfrom∼solarto∼twicesolar
andarangein[a /Fe]from∼0.2to∼0.4.Furthermorethecomparisonofthederivedparameters
atdifferentgalactocentricdistancesshowsthepresenceofnegativemetallicitygradientsfromthe
centeroutwards.
BaryonsinDarkMatterHalos
5-9October2004
Novigrad,Croatia
∗Speaker.
PublishedbySISSA http://pos.sissa.it/
Nearbyearly–typegalaxieswithionizedgas FrancescaAnnibali
1. Introduction
Although Early-Type galaxies (Es)appear asahomogeneous class, characterized bywellde-
fined scaling relations like the Fundamental Plane, the color-magnitude relation and the Mg2-s
relation, much evidence suggests that a secondary episode of star formation has occurred dur-
ing their evolutionary history. Simulations indicate that galaxy collisions, accretion and merging
episodesareimportantfactorsintheevolutionofgalaxyshapes(seee.g. Barnes(1996);Schweizer
(1996)) and can interfere with their passive evolution. This understanding of early-type galaxy
formation hasbeen enhanced bythestudy ofinterstellar matter. Thiscomponent anditsrelevance
inseculargalacticevolutionwaswidelyneglectedinearlystudiesofearly-typegalaxiessincethey P
wereforalongtimeconsidered tobeessentiallydevoidofinterstellar gas. Inthelasttwodecades,
o
however, multi-wavelength observations have changed thispicture andhavedetected thepresence
S
of a multi-phase Inter Stellar Medium (ISM). For what concerns the “warm” ionized component
(
oftheISM,themainionizationmechanism,whichdoesnotseemtobepoweredbystarformation,
remains still uncertain. With the aim of improving the understanding of the nature of the ionized B
gasinearly-type galaxies,Rampazzoetal. (2005)(hereafterR05)havestartedastudyofasample D
of Es showing emission lines in their spectra. The adopted strategy is to investigate the physical
M
conditions oftheionized gas,thepossible ionization mechanisms andtheconnection tothestellar
H
population ofthehostgalaxy.
2
2. Observed properties oftheSample 0
0
The R05 sample contains 50 early–type galaxies. The sample is selected from a compilation
of nearby galaxies (z<5500 km s−1) showing ISM traces in at least one of the following bands: 4
IRAS 100 µm, X-ray, radio, HI and CO, and should be biased towards objects that might be ex- )
pected to have ongoing and recent star formation because of the presence of emission lines. The
0
emission should comes from a combination of active galactic nuclei and star formation regions
5
within the galaxies. The spectra, taken along the major axis, are provided at different concentric
6
regions (apertures) atradii of1.5",2.5",10",r /10, r /8,r /4andr /2,andinfouradjacent regions
e e e e
(gradients)at0≤r≤r /16("nuclear"),r /16≤r≤r /8,r /8≤r≤r /4andr /4≤r≤r /2. Forallthe
e e e e e e e
apertures andgradients R05measuredthe21line-strength indicesoftheoriginalLick-IDSsystem
using the redefined passbands of Trager et al. (1998) (TR98) plus the higher order Balmer lines
introduced by Worthey & Ottaviani (1997) (WO97). The measured indices have been conformed
to the Lick-IDS System following the procedure described in WO97. The correction procedure
include correction forpossible hydrogen emission, correction forvelocity dispersion andtransfor-
mation to the Lick system. For details on the observations and reduction procedure we refer to
R05.
Thepresenceofemissionlinesaffectsthemeasureofsomeline–strengthindices. Inparticular,
the Hb index measure of the underlying stellar population could be contaminated by a significant
infilling due to presence of the Hb emission component. R05 have measured Hb emissions in
theirsamplefollowingtwodifferentapproaches: 1)measuringthe[OIII]emissionandderivingthe
Hb emission through the relation EW(Hb )/EW([OIII]l 5007) =0.7 found by Gonzalez et al.
em
(1993)forhissampleofearly-typegalaxies;2)measuringtheHa emittedfluxandderivingtheHb
emitted flux according to the relation FHb =1/2.86FHa (see e.g. Osterbrock (1989)). In Figure 1
056/2
Nearbyearly–typegalaxieswithionizedgas FrancescaAnnibali
P
o
S
P
(
B o
Figure 1: (left four panels) R05 sample: Hb emission estimate derived from Ha against the emission
D
S
derivedfrom[OIII]forthefourgradients. Thesolidlineistheone-to-onerelation. Opengreencirclesare
M
forgalaxieswithO[III]emissiondetectedunder1s level,bluetrianglesandredfullsquaresbetween1and (
H
2s levelsandabove2s levelrespectively. (rightfourpanels)Fitting ofN[II](l 6548, 6584)andHa lines
B
2
fortworepresentativegalaxies: onewithHa inemission(NGC2749)andthesecondwiththeHa infilling
0 D
(IC2006). LowerpanelsshowtheresidualslinesafterthesubtractionoftheHa lineofthetemplategalaxy
0
NGC1426(dottedlinesintheupperpanels). M
4
H
)
itisshownthecomparison between thetwodifferent Hb emission estimates computed int0hefour 2
gradients.
5 0
6
0
3. SimpleStellar PopulationIndices
4
Following the procedure described in Bressan et al. (1996), to which we refer for details,
)
we have derived line strength indices for SSPs. In brief all indices are constructed by means of a
central band-pass and two pseudo-continuum band-passes on either side of the central band (see 0
Worthey et al. (1994) (W94), TR98). The continuum flux is interpolated between the mid points 5
ofthepseudo-continuum bandpasses. ThedefinitionoftheindexinEWis
6
I =(1−F /F )Dl (3.1)
EW R C
where F and F are the fluxes in the line and in the pseudocontinuum respectively and Dl
R C
is the width of the central band. The integrated indices for SSPs are calculated according to the
following method. Given an isochrone in the HR diagram, we derive for each elementary bin
D log L andD logT thefluxinthecontinuum F fromlibraries ofstellarspectraandtheindexI
L⊙ eff C
fromthefittingfunctions(FFs)ofW94andWO97. FromF andIthefluxinthecentralpass-band
C
F is computed inverting (3.1). After having integrated F and F along the isochrone, the final
R C R
SSPindexI iscomputedfrom(3.1).
SSP
3.1 a −enhancement
Sincethe fitting functions are calibrated onMilky Waystars, they reflect, atsolar metallicity,
the solar element partitions. Nevertheless several studies have shown that elliptical galaxies are
characterizedbysupersolar[a /Fe]. Modelsaccountingfora -enhancedchemicalcompositionhave
been presented by Thomas et al. (2003, 2004) and Tantalo et al. (2004). The methods used are
based on Response Functions (RFs), introduced for the first time by Tripicco & Bell (1995), that
056/3
Nearbyearly–typegalaxieswithionizedgas FrancescaAnnibali
describe the behavior of indices with elements variations. Similarly to Tantalo et al. (2004), we
have derived RFs from model atmospheres computed with the code ATLAS9 for different Z/Z
⊙
(1/50, 1/5, 3/5), T (4250, 4500, 5000, 5750, 6250, 7000 K), logg (2, 4, 4.5, 5) and [a /Fe] (0.0
eff
and0.4),inordertoconstruct SSPmodelswithnon-solar elementpartitions. ForfixedT ,gand
eff
Z,andvarying [a /Fe]from0to0.4,thefractional variation ofthei-thindexis:
I(i,[a /Fe]=0.4)−I(i,[a /Fe]=0)
R = (3.2)
i,0.4
I(i,[a /Fe]=0)
where I(i,[a /Fe]=0.4) and I(i,[a /Fe]=0) are the indices computed on the synthetic stellar spectra
for solar-scaled and a -enhanced compositions respectively. At metallicities higher than Z/Z⊙ = P
3/5, the responses have been derived through linear extrapolation. At this point we computed
o
a -enhanced SSPs with [a /Fe] = 0.4 in the following way. For each elementary bin (D log L ,
L⊙ S
D logT )alongtheisochrone, weappliedacorrectionD ItotheindexIderivedfromtheFFs. The
eff
i-thindexcorrected fortheeffectofa -enhancement is: (
B
I =I +D I =I +R ×I (3.3)
i,0.4 i,0 i,0.4 i,0 (i,0.4,T,g,Z) i,0
D
whereI isthevaluefromtheFFsandR isobtainedbylinearlyinterpolatingtheR
i,0 i,0.4,T,g,Z i,0.4
M
values of(3.2)atthegravity, temperature andmetallicity ofthebin. Oncecomputed thecorrected
H
indexI,theproceduretoobtainthefinalSSPindexisthesameasdescribedatthebeginningofthe
section forsolar-scaled models. MoredetailscanbefoundinAnnibalietal. (2005)inprep. 2
4. Results 0
0
Wehaveanalyzedasampleofearly-typegalaxiesshowingemissionlinesintheirspectra(R05)
bymeansofnewSSPindicesthataccountalsofortheenhancementofthea -elements. Itisknown 4
thatnarrowbandindicesconstituteapowerfultooltodisentangleage,metallicityandenhancement )
effects(seee.g. Worthey(1994), WO97). Wehavedevisedasimplebutrobustalgorithm toderive
0
these parameters from the analysis of the Mgb, [MgFe] and Hb (or Hg ) indices (Annibali et al.
5
(2005) in prep.). The results of this preliminary investigation are shown in Figure 2. From top
6
to bottom we plot the derived ages, metallicities and [a /Fe] ratios as function of central velocity
dispersion s . Black dots, red squares and green triangles correspond respectively to the nuclear
0
region (r ≤r /16), to r /16 ≤ r ≤r /8 and to r /8 ≤ r ≤r /4. Same colors are used for lines which
e e e e e
indicate the fits to the data. The positive trend of the s –Z and s - [a /Fe] relations of early-type
galaxies is reproduced for our sample. The galaxies span a range in metallicity from ∼ solar (Z
∼ 0.02 ) to ∼ twice solar. The [a /Fe] ratios are supersolar and go from ∼ 0.2 to ∼ 0.4 from the
smallest to the largest systems. The shift between the metallicity relations for the plotted regions
indicates negative Z gradients from the center outwards. The broad range of ages derived from
this analysis is at variance with the quite tight relations obtained for the global metallicity and a -
enhancement. This could suggest that the reliability ofthe whole process ofage determination by
means of Hydrogen indices (possibly affected by emission contamination or by a - enhancement)
is still far from being satisfactory. Indeed there are several data with too high ages. However the
spectra are of high quality and the adopted lines are not heavily contaminated by emission, while
effects of enhancement should include a bias instead of dispersion. We thus conclude that this
is an evidence that the formation process of the bulk of the galaxy follows a well defined path
(and timescale), and that the broad range in age is caused by subsequent "rejuvenation" episodes
056/4
Nearbyearly–typegalaxieswithionizedgas FrancescaAnnibali
P
o
S
(
Figure2: Fromtoptobottomweplotthederivedages,Zand[a /Fe]asfunctionofcentralvelocitydisper-
sions . Blackdots,redsquaresandgreentrianglescorrespondtotheregions0≤r≤r /16,r /16≤r≤r /8 B
0 e e e
andre/8≤r≤re/4. Samecolorsareusedforlineswhichindicatethefitstothedata. D
M
involving only a tiny fraction of the galaxy mass (e.g. Bressan et al. (1996)). Thus the "indirect"
H
signature left on the chemical enrichment path may be a more robust estimator of the formation
timescales thanthe"direct"estimatebasedonHlines, markingthesecularevolution ofgalaxies. 2
References 0
[1] Barnes,J.1996inGalaxies: InteractionsandInducedStarFormation,Saas-FeeAdvancedCourse 0
26.LectureNotes1996.SwissSocietyforAstrophysicsandAstronomy,XIV,Springer-Verlag
4
Berlin/Heidelberg,p.275
)
[2] Bressan,A.,Chiosi,C.,&Tantalo,R.1996,A&A,311,425
0
[3] Gonzalez,J.J.1993,Ph.D.thesis,Univ.California,SantaCruz.
5
[4] Osterbrock,D.,1989,inAstrophysicsofPlanetaryNebulaeandActiveGalacticNuclei,University
ScienceBooks 6
[5] RampazzoR.,AnnibaliF.,BressanA.etal.,2005,A&Aaccepted,astro-ph/0412579
[6] Schweizer,F.1996inGalaxies: InteractionsandInducedStarFormation,Saas-FeeAdvanced
Course26.LectureNotes1996.SwissSocietyforAstrophysicsandAstronomy,XIV,Springer-Verlag
Berlin/Heidelberg,p.105
[7] Tantalo,R.,&Chiosi,C.2004,MNRAS,353,917
[8] Thomas,D.,Maraston,C.,&Bender,R.2003,MNRAS,343,279
[9] Thomas,D.,Maraston,C.,&Korn,A.2004,MNRAS,351,L19
[10] Trager,S.C.,Worthey,G.,Faber,S.M.,Burstein,D.,GozalezJ.J.1998,ApJS116,1
[11] Trager,S.C.,Faber,S.M.,Worthey,G.,&González,J.J.2000,AJ,120,165
[12] Tripicco,M.J.,&Bell,R.A.1995,AJ,110,3035
[13] Worthey,G.,Faber,S.M.,Gonzalez,J.J.,Burstein,D.1994,ApJS94,687
[14] Worthey,G.1994,ApJS,95,107
[15] Worthey,G.,Ottaviani,D.L.1997,ApJS111,377
056/5