Stellar populations in local star-forming galaxies

合集下载

Stellar Populations in Nearby Lenticular Galaxies

a r X i v :a s t r o -p h /0512305v 1 12 D e c 2005Based on the observations collected with the 6m telescope(BTA)at the Special Astrophysical Observatory (SAO)of theRussian Academy of Sciences (RAS).Stellar populations in nearby lenticular galaxiesO.K.Sil’chenko Sternberg Astronomical Institute,Moscow,119992Russia and Isaac Newton Institute of Chile,Moscow Branch Electronic mail:olga@sai.msu.su ABSTRACT We have obtained 2D spectral data for a sample of 58nearby S0galaxies with the Multi-Pupil Spectrograph of the 6m telescope of the Special Astrophysical Observatory of the Russian Academy of Sciences.The Lick indices H β,Mgb,and Fe are calculated separately for the nuclei and for the bulges taken as the rings between R =4′′and 7′′;and the luminosity-weighted ages,metallicities,and Mg/Fe ratios of the stellar populations are estimated by confronting the data to SSP models.Four types of galaxy environments are considered:clusters,centers of groups,other places in groups,and ﬁeld.The nuclei are found to be on average slightly younger than the bulges in any types of environments,and the bulges of S0s in sparse environments are younger than those in dense environments.The eﬀect can be partly attributed to the well-known age correlation with the stellar velocity dispersion in early-type galaxies (in our sample the galaxies in sparseenvironements are in average less massive than those in dense environments),but for the most massive S0s,with σ∗=170−220km/s,the age dependence onthe environment is still signiﬁcant at the conﬁdence level of 1.5σ.Subject headings:galaxies:nuclei —galaxies:elliptical and lenticular —galaxies:evolution1.IntroductionIn classical morphological sequence by Hubble (1936)lenticular galaxies occupy inter-mediate position between ellipticals and spirals:they have a smooth and red appearanceas the ellipticals,but also have stellar disks,almost as large as those of the spirals.The most popular hypothesis of S0origin is that of their transformation from the spirals by stop-ping global star formation and removing or consuming remaining gas(Larson et al.1980). In distant,z∼0.5,clusters this transformation is now observed directly:the number of lenticulars in the clusters diminishes strongly with the redshift(Fasano et al.2000),instead one can see‘passive spirals’–red spiral galaxies lacking star formation–at the periphery (‘infalling regions’)of the intermediate-redshift clusters(Goto et al.2003;Yamauchi&Goto 2004).Many theoretical works have been done to explain in detail what physical mecha-nisms may be involved into the process of spiral transformation into the lenticulars:tidally induced collisions of disk gas clouds(Byrd&Valtonen1990),harassment(Moore et al.1996), ram pressure by intercluster medium(Quilis et al.2000),etc.For the S0s in theﬁeld,the scheme of their transformation from the spirals is not so clear,but common view is that some external action like minor merger may produce the necessary eﬀect.By reviewing the various mechanisms of secular evolution which may transform a spiral galaxy into a lenticular one we have noticed that most of them result in gas concentration in the very center of the galaxy,so that a nuclear star formation burst seems inavoidable circumstance of the S0galaxy birth.If to refer to S0statistics in the clusters located between z=0and z≈1,the main epoch of S0formation is z≈0.4−0.5,so the nuclear star formation bursts in the nearby S0s must not be older than5Gyr.Indeed,in my spectral study of the central parts of nearby galaxies in diﬀerent types of environments(Sil’chenko1993)I have found that∼50%of nearby lenticulars have strong absorption lines Hγand Hδin their nuclear spectra so they are of‘E+A’type,as it is presently called and are dominated by intermediate-age stellar population.In this respect the S0s have resembled rather early-type spirals than ellipticals.Here I aim to continue this study,with a larger sample and with panoramic spectral data in order to separate the nuclei and their outskirts(bulges)which is a substantial advantage with respect to aperture spectroscopy.Another crucial point of the present study,and also of a global paradigm of galaxy formation,is environmental inﬂuence.The current hierarchical assembly paradigm predicts a younger age of galaxies in lower density environments–for the most recent simulations see nzoni et al.(2005)or De Lucia et al.(2005).Observational evidences concerning early-type galaxies are controversial:some authorsﬁnd diﬀerences of stellar population ages between the clusters and theﬁeld(Terlevich&Forbes2002;Kuntschner et al.2002;Thomas et al.2005),some authors do notﬁnd any dependence of the stellar population age on environment density(Kochanek et al.2000).In order to check whether the mean ages of the stellar populations depend on environment density monotonously,as the hierarchical paradigm predicts,in this work I consider four types of environments separately:the cluster galaxies,the brightest(central)galaxies of groups,the second-ranked group members,andtheﬁeld galaxies.2.SampleThe sample of lenticular galaxies considered in this work consists of58objects,mostly nearby and bright.It does not pretend to be complete but rather representative.In the LEDA we have found122galaxies in total with the following parameters:−3≤T≤0, v r<3000km/s,B0T<13.0,δ2000.0>0,without bright AGN or intense present star formation in a nucleus;among them40Virgo members.For our sample,from this list we have selected 8Virgo members and42other galaxies–half of the rest.A few galaxies are added to broaden the luminosity range:NGC5574,NGC3065,and NGC7280are fainter than B0T=13.0, NGC80and NGC2911are very luminous but farther from us than40Mpc.Table1lists all the galaxies with some of their characteristics such as morphological type,redshift,and central velocity dispersion.Sorting of the galaxies according to their environment type has been made by using the NOG group catalogue(Giuricin et al.2000); we have only classiﬁed three galaxies belonging to the Ursa Major cluster following Tully et al.(1996).Our sample includes11cluster galaxies,from Virgo and Ursa Major,17 central(the brightest)galaxies of groups with3members and more,18second-ranked group members to which we have added paired galaxies,and12ﬁeld lenticulars which are deﬁned as not mentioned in the NOG catalogue at all.All the galaxies of Table1have been observed with the integral-ﬁeld unit–the Multi-Pupil Fiber/Field Spectrograph(MPFS)1(Afanasiev et al.2001)of the6m telescope of the Special Astrophysical Observatory of the Russian Academy of Sciences between1994and 2005.For these years the instrument was modiﬁed more than once.We started with theﬁeld of view of10′′×12′′,with the spatial element(pupil)size of1.′′3,with the spectral resolution of5˚A,and spectral range less than600˚A.Now we have16′′×16′′,the spatial element(pupil) size of1′′,the spectral resolution of3.5˚A,and spectral range of1500˚ually we observe two spectral ranges,the green one centered ontoλ5000˚A,and the red one centered onto the Hαline.The optical design had been modiﬁed too:two diﬀerent schemes,a TIGER-like one –for the description of the instrumental idea of the TIGER mode of IFU one can see Bacon et al.(1995)–and that withﬁbers,were used before and after1998.We have described in detail23of58lenticulars in our previous papers(see the references in the Table1)where one canﬁnd not only the characteristics of the various versions of the MPFS,but also2D maps of Lick indices and kinematical parameters.Here we consider only two discrete areas of every galaxy–the unresolved nuclei and the wide rings,with R in=4′′and R out=7′′, which we are treating as the‘bulges’.The boundaries of the rings have been selected as a compromise between the seeing limitation(the seeing FWHM are typically2.′′5at the6m telescope)in order to avoid the inﬂuence of the nuclei on the bulge measurements,and the size of ourﬁeld of view which causes incomplete azimuthal coverage at R>7′′.At our limit distance,D=40Mpc,the outer radius of the‘bulge’areas,7′′,corresponds to the linear size of1.35kpc.The nuclei are presented by the integratedﬂuxes over the central spatial elements within the maximum radius of0.1kpc from the centers.main.htmlTable2:The comparison of two independent index determinations with the MPFSnucleus bulge1.57 1.70 3.14 3.221.661.59±0.202.942.95±0.15936 4.644.51±0.034.934.53±0.101.561.55±0.032.102.07±0.051.731.35±0.052.382.47±0.024036 5.564.09±0.135.853.61±0.251.141.10±0.032.972.00±0.140.740.96±0.06 2.652.11±0.074550 3.18 3.133.203.14±0.051.050.862.92 2.101.56 1.242.76 2.127013 3.84 3.323.783.27±0.052.101.54±0.10 2.922.05±0.162.241.65±0.10 2.80 2.237457 2.72 3.372.922.98±0.06Table3:The mean diﬀerences between our indices and the Trager’s et al.data∆+0.07˚A−0.05˚A+0.12˚A±0.06˚A±0.07˚A±0.09˚ANote.—The second line of the table contains the formal errors of the mean oﬀsets of our index system with respect to the Lick oneThe Lick indices Hβ,Mgb,Fe5270,and Fe5335have been measured for the nuclei and for the bulges of all the galaxies;farther we use the composite iron index Fe ≡(Fe5270+Fe5335)/2. During all our observational runs we observed standard stars from Worthey et al.(1994)and calibrated our index system onto the standard Lick one.The measured indices were corrected for the stellar velocity dispersions;we calculated the corrections by artiﬁcial broadening of the spectra of the standard stars.We estimate the typical statistical accuracy in each of three indices(deﬁned by the S/N ratio which has been kept as70-90(per˚A)in the nuclei and∼30at the edges of the frames)as0.1˚A.Some galaxies of the sample have been ob-served twice.In the Table2we show the raw index measurements from two independent observational runs for each of those objects;±accompanying the bulge indices reﬂect partly the index variations along the radii–we average four measurements at four R’s from4′′to7′′for each galaxy and give here the errors of the means.The mean absolute diﬀerence between two independent index measurements is0.20˚A for the nuclei and0.18˚A for the bulges over the Table2.If we analyse three indices separately,we obtain the mean absolute diﬀerences (the rms of the diﬀerences)of0.22˚A(0.29˚A)for Hβ,0.15˚A(0.19˚A)for Mgb,and0.22˚A (0.28˚A)for the composite iron index.These results mean that the accuracy of the Mgb corresponds to our expectations from the S/N statistics,namely,is0.1˚A,and the accuracy of the Hβand Fe is somewhat worse,namely,is0.15˚A.Among our58galaxies,28objects have Lick index measurements through the central aperture2′′×4′′by Trager et al.(1998). The results of the comparison of these standartized Lick indices with our measurements for the nuclei are presented in Table3and in Fig.1.The smallest scatter is found for Hβand the largest one–for Fe ,that is consistent with the fact the among the four indices,Hβ, Mgb,Fe5270,and Fe5335,the errors quoted by Trager et al.(1998)are the smallest for Hβ(0.24˚A on average over the common list)and the largest–for Fe5335(0.34˚A on average over the common list).In general,our index system does not deviate from the standard Lick one in any systematic way,so we can determine the stellar population properties by confronting our indices to evolutionary synthesis models.–14–3.Stellar population properties in the nuclei and the bulges of S0sTables4and5contain the measured Lick indices Hβ,Mgb,and Fe ≡(Fe5270+ Fe5335)/2for the nuclei and for the bulges correspondingly,as well as the parameters of the stellar population–luminosity-weighted age and metallicity–determined with these indices as described below.Some galaxies have measurements only for the nuclei or only for the bulges due to various reasons–for example,in NGC5866the nucleus is completely obscured by dust and in NGC676the bulge measurements are severely contaminated by a bright star projected at5′′from the nucleus.For the indices presented here,there are models based on evolutionary synthesis of simple(one-age,one-metallicity)stellar populations–see e.g. Worthey(1994).These models allow to estimate the luminosity-weighted mean metallicities and the ages of the stellar populations by confronting the hydrogen-line index Hβto any metal-line index.We are also going to consider the duration of the last major star-forming episode by confronting Fe to Mgb.Chemical evolution models,see e.g.Matteucci(1994), show that because of the diﬀerence in the timescales of iron and magnesium production by a stellar generation,the solar Mg/Fe abundance ratio can be obtained only by very continuous star formation,and brief star formation bursts,withτ≤0.1Gyr,would give signiﬁcant magnesium overabundance,up to[Mg/Fe]=+0.3−+0.4.In this work we use recent models by Thomas et al.(2003)because these models are calculated for several values of[Mg/Fe]:they allow to estimate Mg/Fe ratios of the stellar populations from Mgb and Fe ≡(Fe5270+Fe5335)/2measurements.Figure2presents the Fe vs Mgb diagrams for the bulges and Fig.3–the similar diagrams for the nuclei,for all four types of environments.For some galaxies(e.g.NGC2655 and NGC2911)where the N Iλ5199emission is signiﬁcant,the Mgb indices are corrected from this emission line according to the prescription of Goudfrooij&Emsellem(1996).The model sequences for[Mg/Fe]=0.0,+0.3,and+0.5are well separated on the diagrams Fe vs Mgb,so we can estimate the mean Mg/Fe ratios‘by eye’.Surprisingly,the bulges of the group central galaxies diﬀer from those of the second-ranked group members:the former have the mean[Mg/Fe]≈+0.2,and the latter–+0.1.As by the deﬁnition the second-rank group galaxies are less luminous than the central ones,this diﬀerence may be attributed not to the environment density,but to the galaxy mass eﬀect,at theﬁrst glance.To check this,in Fig.4we have plotted the bulges only for the galaxies within the narrow stellar velocity dispersion range,σ∗=145−215km/s–in thisσ∗range the central and second-rank group members of our sample have the same meanσ∗of172km/s;still the diﬀerence between the central group galaxies and the second-rank members persists in Fig.4.This tendency of the S0s in the centers of groups to resemble more the cluster lenticulars,and of the second-rank group members and the paired galaxies to be like theﬁeld S0s,is in general conﬁrmed by the nuclei distribution in the Fe vs Mgb diagrams(Fig.3),though there are–15–more‘outliers’among the nuclei:evidently,the evolution of nuclear stellar populations bears more individual features than that of the bulges.To break the age-metallicity degeneracy and to determine simultaneously the mean luminosity-weighted ages and the metallicities of the stellar populations,we confront the Hβindices to the combined metal-line index[MgFe]≡(Mgb Fe )1/2–by plotting our data together with the models of Thomas et al.(2003);earlier we have assured that this diagram is insensitive to the Mg/Fe ratio.However we have one serious problem here:the absorption-line index Hβmay be contaminated by emission,especially in the nuclear spectra.To correct from the emission the Hβindices which we have measured we have used data on equivalent widths of Hαemission lines because Hαemission lines are always much stronger than Hβemission lines and because an Hαabsorption line is not deeper than an Hβabsorption line in spectra of stellar populations of any age while in intermediate-age population spectra it is much shallower.The emission-line intensity ratio,Hα/Hβ,has been studied well both empirically and theoretically.The minimum value of this ratio,2.5,is known for the case of radiative excitation by young stars(Burgess1958).For other types of excitation this ratio is higher.We have no pure H II-type nuclei in our sample,so here we use the formula EW(Hβemis)=0.25EW(Hαemis):this mean relation is obtained by Stasinska&Sodr´e(2001) for a quite heterogeneous sample of nearby emission-line galaxies.The data on EW(Hαemis) for the nuclei we take mainly from Ho et al.(1997).The bulge Hβindices were corrected from the emission by using Hαequivalent widths obtained with the red MPFS spectra for about a half of the sample(28objects,see the Table1).Unlike Ho et al.(1997)who obtained EW(Hαemis)by subtracting a pure absorption-line template from the observed spectra,we applied a multicomponent Gauss-analysis to the combinations of the Hαabsorption and emission lines which was eﬀective due to mostly diﬀerent velocity dispersions of stars and gas clouds in the galaxies under consideration.From the rest,16galaxies have negligible emission lines in the bulge spectra(EW([O III])≤0.3˚A,see the Table5),and for the others the age estimates obtained by using the Hβindices’corrected through the Hα’(Table5) are indeed only upper limits.To correct in some way ALL the bulge spectra,we have used also the wide-known approach which involves the[O III]λ5007emission line;Trager et al. (2000)recommend to use the statistical correction∆Hβ=0.6EW([OIII]λ5007)though they note that individual ratios Hβ/[O III]may vary between0.33and1.25within their sample of elliptical galaxies.In Fig.5we compare the corrections obtained by two diﬀerent ways for the nuclei.If we exclude two galaxies with extremely strong emission in the centers–NGC4138 and NGC7743–statistically the two types of the corrections are indistinguishable;however the accuracy of[O III]measuring is not very high due to strong underlying absorption lines of Ti I,and the weak emission lines[O III]with the equivalent widths of EW≤0.3˚A are evidently artifacts.To summarize this analysis,we conclude that while for mutual–16–comparisons of the age distributions we must take only the age estimates corrected through the[O III]because this correction can be made for all galaxies of the sample,for the individual galaxies having the red spectra the estimates made with the Hβindices corrected through the Hαare more reliable due to the facts that the Hαemission is stronger and that the ratio of the Balmer emission lines depends only on the excitation mechanism unlike the ratio of Hβto[O III]which depends also on the metallicity of the gas.Figure6presents the diagrams Hβvs[MgFe]for the nuclei(top)and for the bulges (bottom)of the galaxies of all types of environments with their Hβindices corrected through the Hαto the left and with their Hβindices corrected through the[O III]to the right, correspondingly.By inspecting these diagrams,we determine the ages and the metallicities ‘by eye’that provides an accuracy of∼0.1dex in metallicity and1Gyr for the ages less than 8Gyr and∼2Gyr for older stellar systems which match our accuracy of the Lick indices. Directly in the diagrams one can see that the range of the ages of the nuclei is very wide: they may be as young as1Gyr old and as old as15Gyr old.The bulges are on average older than the nuclei,and in the bottom plots one can see a segregation of the galaxies according to their type of environment:most the bulges of the group centers and the cluster galaxies are older than5Gyr,whereas some of the group members and theﬁeld lenticulars have the bulges as young as2-3Gyr old.The metallicity ranges seem to be similar for the bulges in all types of environments:their[Z/H]are conﬁned between∼−0.3and∼+0.4.Byﬁtting formally the metallicity distributions by Gaussians,we obtain the mean metallicity for the bulges in dense environments to be–0.04and that for the bulges in sparse environments to be–0.13,with the similar rms of0.5dex.The nuclei seem to be on average more metal-rich: only three nuclei in the galaxies of sparse environments have the metallicity less than the solar.Kuntschner et al.(2002)have already reported the diﬀerence between the stellar popu-lation characteristics of the early-type galaxies in the dense and sparse environments.Their measurements were aperture spectroscopy,and their samples were the Fornax cluster as an example of dense environments and galaxies without more than2neighbors within the search radius of1.3Mpc as an example of sparse environments–the latter sample is probably close to our combination of theﬁeld plus paired galaxies.They have found that the E/S0galaxies in sparse environments are younger than the E/S0galaxies in the cluster by2-3Gyr–and our result for the S0s is quite the same.But they have also found the anti-correlation between the age and metallicity,the younger galaxies in sparse environments being on average more metal-rich(by0.2dex)than the older galaxies in the cluster;while if we see any metallicity diﬀerence,it should be in opposite sense.In Fig.7we plot cumulative distributions of the ages:the number of galaxies not older–17–than T versus log T(in Gyr).We have united the samples of the brightest group S0s and the cluster galaxies into a‘dense environment’sample,and the group second-ranked members and theﬁeld S0s–into a‘sparse environment’sample.The eﬀect of environments is seen both for the nuclei and for the bulges:in sparse environments the stellar populations are, on average,younger.The estimates of the median ages are the following:3.7and6Gyr for the nuclei of the galaxies in sparse and dense environments,correspondingly,and4.8and 8.3Gyr for the bulges.4.DiscussionIt is a little bit surprising that according to my results,the‘dense’type of environment must be ascribed not only to the clusters but also to the centers of groups:theﬁrst-ranked and the second-ranked S0galaxies of the groups have very diﬀerent properties of their central stellar population.However,this conclusion is close to the recentﬁnding by Proctor et al. (2004)that the early-type galaxies of Hickson compact groups resemble more the cluster galaxies than theﬁeld ones.I think it gives us a hint that the dynamical eﬀect of close neighbors may play the main role in evolution rate,and not the mass of the common dark halo.–18–Table6:The mean ages of the bulges withinﬁxed stellar velocity dispersion ranges Range ofσ∗,km/s N gal T ,Gyr Its rms105–14564.5±1.0 2.386.5±1.0 2.6185–22558.6±1.9 3.9–19–Recently some evidences have been published(Caldwell et al.2003;Nelan et al.2005) that the ages of the stellar populations in early-type galaxies are correlated with the central stellar velocity dispersion.In our sample,the galaxies in dense environments are on average more massive than those in sparse environments so one may suggest that the age diﬀerence found above may be due to the mass diﬀerence and not to the environment inﬂuence.To check this eﬀect,I have plotted the bulge age estimates versus the central stellar velocity dispersion in Fig.8.Indeed,the correlation is present implying that the more massive bulges are older;the slope of the regression log T vs logσ∗,0is1.76±0.65for the dense environments and1.30±0.43for the sparse ones with the correlation coeﬃcients of0.53and 0.55,correspondingly.Following Caldwell et al.(2003),we have calculated the mean ages of the bulges within narrow ranges of stellar velocity dispersion(when the age estimate has only the low limit of15Gyr,I have ascribed the value of16Gyr to it).These estimates are given in Table6–please compare them with those in Caldwell et al.(2003),7.4Gyr,rms4.2 Gyr,in the range ofσ∗=100−160km/s,and9.9Gyr,rms4.2Gyr,in the range ofσ∗>160 km/s.One can see from Table6that the ages of the bulges are the same in diﬀerent types of environments for the lower bins ofσ∗,105–145and145-185km/s;but in the highest bin, 185–225km/s,the ages are dramatically diﬀerent,the massive bulges in dense environments being much older than the massive bulges in sparse environments.By inspecting Fig.8,we notice that the separation between the bulges in diﬀerent types of environments starts from aboutσ∗=170km/s.Taking7galaxies in dense environments and7galaxies in sparse environments with theσ∗in the range of170–215km/s,we obtain T =9.7±1.3Gyr,rms 3.2Gyr,for the former and T =6.6±1.5Gyr,rms3.7Gyr,for the latter subsample;so the diﬀerence is3.1±2.0Gyr.The application of the Student T-statistics to this double subsample of the massive bulges shows that the mean age of the massive bulges in dense environments is larger than the mean age of the massive bulges in sparse environments with the probability higher than0.85(the hypothesis of T dense≤ T sparse is rejected at the signiﬁcance level of0.14).5.ConclusionsBy considering the stellar population properties in the nuclei and the bulges of the nearby lenticular galaxies in the various types of environments,I have found certain diﬀer-ences between the nuclei and the bulges as well as between the galaxies in dense and sparse environments.The nuclei are on average younger than the bulges in any types of environ-ments,and both the nuclei and the bulges of S0s in sparse environments are younger than those in dense environments.The results of the consideration of the Mg/Fe ratios suggest that the main star formation epoch may be more brief in the centers of the galaxies in dense–20–environments.I am grateful to the astronomers of the Special Astrophysical Observatory of RAS V.L. Afanasiev,A.N.Burenkov,V.V.Vlasyuk,S.N.Dodonov,and A.V.Moiseev for supporting the MPFS observations at the6m telescope.The6m telescope is operated under theﬁnan-cial support of Science Ministry of Russia(registration number01-43);we thank also the Programme Committee of the6m telescope for allocating the observational time.During the data analysis we have used the Lyon-Meudon Extragalactic Database(LEDA)supplied by the LEDA team at the CRAL-Observatoire de Lyon(France)and the NASA/IPAC Ex-tragalactic Database(NED)which is operated by the Jet Propulsion Laboratory,California Institute of Technology,under contract with the National Aeronautics and Space Adminis-tration.The study of the young nuclei in lenticular galaxies was supported by the grant of the Russian Foundation for Basic Researches no.01-02-16767.REFERENCESAfanasiev,V.L.,Sil’chenko,O.K.2000,AJ,119,126Afanasiev,V.L.,Sil’chenko,O.K.2002,AJ,124,706Afanasiev,V.L.,Dodonov,S.N.,Moiseev,A.V.,2001,In:Stellar dynamics:from classic to modern/Eds.Osipkov L.P.and Nikiforov I.I.,Saint Petersburg Univ.press,103 Bacon,R.,Adam,G.,Baranne,A.,Courtes,G.,Bubet,D.,et al.1995,A&AS,113,347 Burgess,A.,1958,MNRAS,118,477Byrd,G.,Valtonen,M.,1990,ApJ,350,89Caldwell,N.,Rose,J.A.,Concannon,K.D.2003,AJ,125,2891De Lucia,G.,Springel,V.,White,S.D.M.,Croton,D.,Kauﬀmann,G.2005,MNRAS,in press,astro-ph/0509725Fasano,G.,Poggianti,B.M.,Couch,W.J.,Bettoni,D.,Kjaergaard,P.,Moles,M.,2000, ApJ,542,673Giuricin,G.,Marinoni,C.,Ceriani,L.,Pisani,A.2000,ApJ,543,178Goto,T.,Okamura,S.,Sekiguchi,M.,Bernardi,M.,Brinkmann,J.,et al.2003,PASJ,55, 757Goudfrooij,P.,Emsellem,E.1996,A&A,306,L45Ho,L.C.,Filippenko,A.,Sargent,W.L.1997,ApJS,112,315Hubble,E.P.,1936,Realm of the Nebulae,Yale Univ.PressKochanek,C.S.,Falco,E.E.,Impey,C.D.,Leh´a r,J.,McLeod,B.A.,et al.,2000,ApJ,543, 131,2000Kuntschner,H.,Smith,R.J.,Colless,M.,Davies,R.L.,Kaldare,R.,Vazdekis,A.2002, MNRAS,337,172Lanzoni,B.,Guiderdoni,B.,Mamon,G.A.,Devriendt,J.,Hatton,S.2005,MNRAS,361, 369Larson,R.B.,Tinsley,B.M.,Caldwell,C.N.1980,ApJ,237,692Matteucci,F.1994,A&A,288,57Moore,B.,Katz,N.,Lake,G.,Dressler,A.,Oemler,A.,Jr.,1996,Nature,379,613 Nelan,J.E.,Smith,R.J.,Hudson,M.J.,Wegner,G.A.,Lucey,J.R.,et al.2005,ApJ,632, 137Proctor,R.N.,Forbes,D.A.,Hau,G.K.T.,Beasley,M.A.,De Silva,et al.,2004,MNRAS, 349,1381Quilis,V.,Moore,B.,Bower,R.,2000,Science,288,1617Sil’chenko,O.K.1993,Pis’ma v AZh,19,701Sil’chenko,O.K.1998,A&A,330,412Sil’chenko,O.K.1999,AJ,117,2725Sil’chenko,O.K.2000,AJ,120,741Sil’chenko,O.K.,Vlasyuk V.V.2001,Pis’ma v AZh,27,19Sil’chenko,O.K.,Afanasiev,V.L.2002,A&A,385,1Sil’chenko,O.K.,Afanasiev,V.L.2004,AJ,127,2641Sil’chenko,O.K.,Afanasiev V.L.,Chavushyan,V.H.,Valdes,J.R.2002,ApJ,577,668 Sil’chenko,O.K.,Moiseev,A.V.,Afanasiev,V.L.,Chavushyan,V.H.,Valdes,J.R.2003,ApJ, 591,185Sil’chenko,O.K.,Koposov,S.E.,Vlasyuk,V.V.,Spiridonova,O.I.2003,Astronomy Reports, 47,88Stasinska,G.,Sodr´e,I.,Jr.2001,A&A,374,919Terlevich,A.I.,Forbes,D.A.2002,MNRAS,330,547Thomas,D.,Maraston,C.,Bender,R.2003,MNRAS,339,897Thomas,D.,Maraston,C.,Bender,R.,Mendes de Oliveira,C.,2005,ApJ,621,673 Trager,S.C.,Worthey,G.,Faber,S.M.,Burstein,D.,Gonz`a lez,J.J.1998,ApJS,116,1 Trager,S.C.,Faber,S.M.,Worthey,G.,Gonz`a lez,J.J.2000,AJ,119,1645。

Stellar populations in the nuclear regions of nearby radiogalaxies

a r X i v :a s t r o -p h /0103059v 1 4 M a r 2001Mon.Not.R.Astron.Soc.000,000–000(0000)Printed 1February 2008(MN L A T E X style ﬁle v1.4)Stellar populations in the nuclear regions of nearbyradiogalaxiesItziar Aretxaga 1,Elena Terlevich 1⋆,Roberto J.Terlevich 2†,Garret Cotter 3,´Angeles I.D´ıaz 41Instituto Nacional de Astrof´ısica,´Optica y Electr´o nica,Apdo.Postal 25y 216,72000Puebla,Pue.,Mexico 2Institute of Astronomy,Madingley Road,Cambridge CB30HA,U.K.3Cavendish Laboratory,Univ.of Cambridge,Madingley Road,Cambridge CB30HE,U.K.4Dept.F´ısica Te´o rica C-XI,Univ.Aut´o moma de Madrid,Cantoblanco,Madrid,Spain.1February 2008ABSTRACTWe present optical spectra of the nuclei of seven luminous (P 178MHz >∼1025W Hz −1Sr −1)nearby (z <0.08)radiogalaxies,which mostly correspond to the FR II class.In two cases,Hydra A and 3C 285,the Balmer and λ4000˚A break indices constrain the spectral types and luminosity classes of the stars involved,re-vealing that the blue spectra are dominated by blue supergiant and/or giant stars.The ages derived for the last burst of star formation in Hydra A are between 7and 40Myr,and in 3C 285about 10Myr.The rest of the narrow-line radiogalaxies (four)have λ4000˚A break and metallic indices consistent with those of elliptical galaxies.The only broad-line radiogalaxy in our sample,3C 382,has a strong featureless blue continuum and broad emission lines that dilute the underlying blue stellar spectra.We are able to detect the Ca II triplet in absorption in the seven objects,with good quality data for only four of them.The strengths of the absorptions are similar to those found in normal elliptical galaxies,but these values are both consistent with single stellar populations of ages as derived from the Balmer absorption and break strengths,and,also,with mixed young+old populations.Key words:galaxies:active –galaxies:starbursts –galaxies:stellar content1.INTRODUCTIONIn recent years new evidence that star formation plays an important role in Active Galactic Nuclei (AGN)has been gathered:•The presence of strong Ca II λλ8494,8542,8662˚A triplet (CaT)absorptions in a large sample of Seyfert 2nuclei has provided direct evidence for a population of red supergiant stars that dominates the near-IR light (Terlevich,D´ıaz &Terlevich 1990).The values found in Seyfert 1nuclei are also consistent with this idea if the dilution produced by a nuclear non-stellar source is taken into account (Terlevich,D´ıaz &Terlevich 1990,Jim´e nez-Benito et al.2000).The high mass-to-light ratios L (1.6µm)/M inferred in Seyfert 2nuclei also indicate that red supergiants dominate the nuclear light⋆Visiting Fellow at IoA,UK†Visiting Professor at INAOE,Mexico (Oliva et al.1995),but a similar conclusion does not holdfor Seyfert 1nuclei.•The absence of broad emission lines in the direct optical spectra of Seyfert 2nuclei which show broad lines in polar-ized light can be understood only if there is an additional central source of continuum,most probably blue stars (Cid Fernandes &Terlevich 1995,Heckman et al.1995).This con-clusion is further supported by the detection of polarization levels which are lower in the continuum than in the broad lines (Miller &Goodrich 1990,Tran,Miller &Kay 1992).•Hubble Space Telescope imaging of the Seyfert Mrk 447reveals that the central UV light arises in a resolved region of a few hundred pc,in which prominent CaT absorption and broad He II λ4686˚A emission lines reveal the red super-giant and Wolf Rayet stars of a powerful starburst.The stars dominate the UV to near-IR light directly received from the nucleus (Heckman et al.1997).At least 50per cent of the light emitted by the nucleus is stellar,as a conservative es-timate.Mrk 447is not a rare case:a large sample of nearby bright Seyfert 2s and LINERs show similar resolved star-c0000RAS2I.Aretxaga,E.Terlevich,R.Terlevich,G.Cotter,A.I.D´ıazburst nuclei of80to a few hundred pc in size(Colina et al. 1997,Gonz´a lez-Delgado et al.1998,Maoz et al.1995,1998), with some of the Seyfert2containing dominant Wolf-Rayet populations(Kunth&Contini1999,Cid Fernandes et al. 1999).A starburst–AGN connection has been proposed in at least three scenarios:starbursts giving birth to massive black holes(e.g.Scoville&Norman1988);black holes being fed by surrounding stellar clusters(e.g.Perry&Dyson1985,Peter-son1992);and also pure starbursts without black holes(e.g. Terlevich&Melnick1985,Terlevich et al.1992).The evi-dence for starbursts in Seyfert nuclei strongly supports some kind of connection.However,it is still to be demonstrated that starbursts play a key role in all kinds of AGN.One of the most stringent tests to assess if all AGN have associated enhanced nuclear star formation is the case of lobe-dominated radio-sources,whose host galaxies have relatively red colours when compared to other AGN vari-eties.In this paper we address the stellar content associated with the active nuclei of a sample of FR II radiogalaxies, the most luminous class of radiogalaxies(Fanaroﬀ&Ri-ley1974)which possess the most powerful central engines and radio-jets(Rawlings&Saunders1991).The presence of extended collimated radio-jets,which fuel the extended radio structure over>∼108yr,strongly suggests the exis-tence of a supermassive accreting black hole in the nuclei of these radiogalaxies.This test addresses the question of whether AGN that involve conspicuous black holes and ac-cretion processes also contain enhanced star formation.In section2we introduce the sample and detail the data acquisition and reduction processes.In section3we provide continuum and line measurements of the most prominent features of the optical spectra of the radiogalaxies.In sec-tion4we discuss the main stellar populations responsible for the absorption and continuum spectra.In section5we oﬀer notes on individual objects.A sumary of the main con-clussions from this work is presented in section6.2.DATA ACQUISITION AND REDUCTION Our sample of radiogalaxies was extracted from the3CRR catalogue(Laing,Riley and Longair1983)with the only selection criteria being edge-brightened morphology,which deﬁnes the FR II class of radiogalaxies(Fanaroﬀ&Ri-ley1974),and redshift z<0.08.This last condition was imposed in order to be able to observe the redshifted CaT at wavelengths shorter thanλ9300˚A,where the atmo-spheric bands are prominent.Six out of a complete sam-ple of ten FR II radiogalaxies that fulﬁll these require-ments were randomly chosen.In addition to this sub-sample of FR IIs,we observed the unusually luminous FR I ra-diogalaxy Hydra A(3C218).This has a radio luminosity of P178MHz=2.2×1026W Hz−1Sr−1,which is an order of magnitude above the typical FR I/FR II dividing luminos-ity.Spectroscopic observations of a total of seven radio-galaxies,one normal elliptical galaxy to serve as reference andﬁve K III stars to serve as velocity calibrators were per-formed using the double-arm spectrograph ISIS mounted in the Cassegrain Focus of the4.2m William Herschel Tele-scope‡in La Palma during two observing runs,in1997 November7–8and1998February19–20.Theﬁrst run was photometric but the second was not,being partially cloudy on the20th.The seeing,as measured from the spatial di-mension of spectrophotometric stars,was between0.7and0.8arcsec throughout the nights.A slit width of1.2arcsec centered on the nucleus of galaxies and stars was used.We oriented the slit along the radio-axis for all radiogalaxies,except for Hydra A,for which the orientation was perpendicular to the radio-axis.An R300B grating centered atλ4500˚A with a2148x4200 pixel EEV CCD and an R316R grating centered atλ8800˚A with a1024x1024pixel TEK CCD were used in the1998 run.The projected area on these chips is0.2arcsec/pixel and0.36arcsec/pixel respectively.This conﬁguration pro-vides the spectral resolution necessary to resolve the Mg b and CaT features and,at the same time,oﬀers a wide spec-tral span:λ3350˚A—λ6000˚A at5.1˚A resolution in the blue andλ7900˚A—λ9400˚A at3.5˚A resolution in the red.In the 1997run,in which we assessed the viability of the project, we used the R600B and R600R gratings instead.This setup covers theλ3810˚A—λ5420˚A andλ8510˚A—λ9320˚A range in the blue and red arm,at2.6and1.7˚A resolution respectively. Just one radiogalaxy(DA240)was observed with this alter-native setup.The dichroics5700and6100were used in1997 November and1998February,and in both runs we used a ﬁlter to avoid second order contamination in the spectra.We obtainedﬂux standards(HZ44and G191-B2B)for the four nights and gratings,except in1998Feb20,when we were unable to acquire the red spectrum of the correspond-ing standard due to a technical failure.One calibration lamp CuAr+CuNe exposure per spectral region and telescope po-sition was also obtained for all objects.The total integration times for the radiogalaxies(from1 to3hr)were split into time intervals of about1200or1800s in order to diminish the eﬀect of cosmic rays on individual frames and allow to take lampﬂat-ﬁelds with the red arm of the spectrograph between science exposures.The TEK CCD has a variable fringing pattern at the wavelengths of interest, such that the variations are correlated with the position at which the telescope is pointing.Sinceﬂat-ﬁelding is crucial for the reddest wavelengths,where the sky lines are most prominent,after every exposure of20to30min we acquired aﬂat-ﬁeld in the same position of the telescope as the one for which the galaxies were being observed.We followed this procedure with all galaxies except with DA240.The same procedure was also used in the case of the elliptical galaxy, splitting its total integration time in two.Table1summarizes the journal of observations,where column1gives the name of the object;column2the radio-power at178MHz;column3the redshift;column4the integrated V magnitude of the galaxy;column5identiﬁes whether the object is a radiogalaxy(RG),a normal elliptical (E)or a star(S);column6gives the date of the beginning of the night in which the observations were carried out;col-umn7the position angle(PA)of the slit;column8the total ‡The William Herschel Telescope is operated on the island of La Palma by the Isaac Newton Group in the Spanish Observatorio del Roque de los Muchachos of the Instituto de Astrof´ısica de Canariasc 0000RAS,MNRAS000,000–000Stellar populations in the nuclear regions of nearby radiogalaxies3exposure time;column9the grating used;and column10 the corresponding linear size to1arcsec at the redshift of the galaxies(for H0=50km s−1Mpc−1).The data for the radiogalaxies were extracted from the3C Atlas(Leahy,Bri-dle&Strom,/atlas/)and for the host galaxy of Hydra A from the3CR Catalogue(de Vaucouleurs et al.1991).The data were reduced using the IRAF software pack-age.The frames wereﬁrst bias subtracted and thenﬂat-ﬁeld corrected.In the case of the red arm spectra,the diﬀerent ﬂats obtained for a single object were combined when no sig-niﬁcant diﬀerences were detected between them.However,in several cases the fringing pattern shifted positions that ac-counted for diﬀerences of up to20per cent.In these cases we corrected each science frame with theﬂat-ﬁeld acquired immediately before and/or afterwards.Close inspection of the faintest levels of theﬂat-ﬁelded frames showed that the fringing had been successfully eliminated.Wavelength and ﬂux calibration were then performed,and the sky was sub-tracted by using the outermost parts of the slit.The atmospheric bands,mainly water absorption at λ8920˚A—λ9400˚A,aﬀect the redshifted CaT region of sev-eral radiogalaxies.The bands have been extracted using a template constructed from the stellar spectra obtained each night.The template was built averaging the normalizedﬂux of spectrophotometric and velocity standard stars,once the stellar absorption lines had been removed.The atmospheric bands were eliminated from the spectra of the galaxies di-viding by theﬂux-scaled template.This reduces the S/N of the region under consideration,especially since the bands are variable in time and one of our observing nights was par-tially cloudy.However,the technique allows the detection of the stellar atmospheric features.The CaT of the elliptical galaxy is not aﬀected by atmospheric absorption.Figure1shows the line spectrum of the sky and,as an example,the atmospheric absorption template of1998 Feb19.Water-band correction proved to be critical for the detection of the CaT lines when the atmospheric conditions were most adverse.Figure2shows extractions of the nuclear2arcsec of the spectra of the galaxies.This corresponds to844to2020pc for the radiogalaxies,and98pc for the normal elliptical galaxy.3.LINE AND CONTINUUMMEASUREMENTS3.1.CaT indexThe CaT was detected in all of the objects,although in three cases(3C285,3C382and4C73.08)it was totally or par-tially aﬀected by residuals left by the atmospheric band cor-rections and the measurement of its strength was thus pre-cluded.For the remaining cases,the strength was measured in the rest-frame of the galaxies against a pseudo-continuum, following the deﬁnition of the CaT index of D´ıaz,Ter-levich&Terlevich(1989).In Hydra A,3C285and3C382, the red continuum band is seriously aﬀected by residuals left from the atmospheric absorption removal.We deﬁned two alternative continuum bands,8575˚A<λ<8585˚A and 8730˚A<λ<8740˚A,that substitute the red-most contin-uum window of the CaT index.We checked this new deﬁni-tion against the original one in the elliptical galaxy,which doesn’t have residuals in its continuum bands,and the agree-ment between the two systems was good within5per cent.Velocity dispersions were measured by cross-correlating the galaxy spectra with the stellar spectra obtained with the same setup.The errors in the velocity dispersions calculated in this way were less than10per cent.A high velocity dispersion tends to decrease the mea-sured values of indices based on EW measurements.The CaT index has to be corrected from broadening of the ab-sorption lines by the corresponding velocity dispersion.In order to calculate the correction we convolved stellar pro-ﬁles with gaussian functions of increasing width,and mea-sured the CaT index in them.A good description of the cor-rection found for our data is given by the functional form ∆EW(˚A)=(σ(km s−1)−100)/200.The corrections were applied to the values measured in the galaxies,and con-verted into unbroadened indices.The values of velocity dispersions(σ),uncorrected EW (CaT u)and corrected EW(CaT),are listed in Table2. 3.2.λ4000˚A and Balmer Break indicesStellar populations can be dated through the measurement of theλ4000˚A or Balmer breaks.In intermediate to old pop-ulations the discontinuity atλ4000˚A results from a combi-nation of the accumulation of the Balmer lines towards the limit of the Balmer absorption continuum atλ3646˚A(the Balmer break)and the increase in stellar opacity caused by metal lines shortwards ofλ4000˚A.Table3lists the values of theλ4000˚A break index,∆4000˚A,measured in the spectra of the6narrow-line ra-diogalaxies and the elliptical galaxy in our sample.This ex-cludes3C382,which has a spectrum dominated by a strong blue continuum and broad-emission lines,and shows very weak stellar atmospheric features and no break.We adopted the deﬁnition given by Hamilton(1985),which quantiﬁes the ratio of the averageﬂux-level of two broad bands,one cov-ering the break(3750-3950)and one bluewards of the break (4050-4250).Both bands contain strong metallic and Balmer absorption lines in the case of normal galaxies.In active galaxies,the measurement can be contaminated by emission of[Ne III]λ3869˚A,which in our case is weak.The contami-nation by high-order Balmer lines in emission is negligible. The net eﬀect of emission contamination is to decrease the Balmer break index.In the radio-galaxies,we have estimated this eﬀect by interpolating the continuum levels below the [Ne III]emission,and we estimate that the ratio can be af-fected by6per cent at worst,in the case of3C192,and by less than3per cent for the rest of the objects.Table3lists emission-devoid indices.Hydra A and3C285have spectra which are much bluer than those of normal elliptical galaxies.In order to quantify better the strength of the break and the ages of the popula-tions derived,we have performed a bulge subtraction using as template the spectrum of NGC4374,scaled to eliminate the G-band absorption of the radiogalaxies.Since the ve-locity dispersion of the stars in NGC4374and in the ra-diogalaxies are comparable inside the spectral resolution of our data,no further corrections were needed.The G-band absorption is prominent in stars of spectral types later thanc 0000RAS,MNRAS000,000–0004I.Aretxaga,E.Terlevich,R.Terlevich,G.Cotter,A.I.D´ıazF5and it is especially strong in types K.NGC4374is a normal elliptical galaxy,with a spectral shape which com-pares well with those of other normal ellipticals in the spec-trophotometric atlas of galaxies of Kennicutt(1992).Thus, by removing a scaled template of NGC4374,we are isolat-ing the most massive stars(M>∼1M⊙)in the composite stellar population of the radiogalaxies.Figure3shows the bulge subtractions obtained on these two radiogalaxies.We measured on the bulge-subtracted spectra∆4000˚A and also the Balmer break index as deﬁned by the classi-cal Dλ1method of stellar classiﬁcation designed by Barbier and Chalonge(Barbier1955,Chalonge1956,see Str¨o mgren 1963).The latter quantiﬁes the Balmer discontinuity in terms of the logarithmic diﬀerence of the continuum levels (D)and the eﬀective position of the break(λ1).The method places a pseudo-continuum on top of the higher order terms of the Balmer series in order to measure the eﬀective posi-tion of the discontinuity.Figure4shows the placement of continua,pseudo-continua and the measurements of D and λ1for an A2I star from the stellar library of Jacoby,Hunter &Christian(1984).The functional dependences on the ef-fective temperature and gravity of the stars are suﬃciently diﬀerent for D andλ1to satisfy a two-dimensional classiﬁ-cation.The Dλ1method could only be reliably applied in the cases of Hydra A and3C285.For the other radiogalaxies, the bulge-subtractions led to results that did not allow the identiﬁcation of the absorption features and/or the break in an unambiguous way due to the resulting poor S/N.Fig-ure5shows the Dλ1measurements performed on the bulge-subtracted spectra of Hydra A and3C285.We have placed diﬀerent continuum levels to estimate the maximum range of acceptable parameters of the stellar populations that are involved.Table3lists theλ4000˚A and Balmer break indices mea-sured in both the bulge-subtracted and the original spectra of the radiogalaxies.3.3.Lick indicesThe presence of prominent Balmer absorption lines,from Hγup to H12λ3750˚A,is one of the most remarkable features of the blue spectra of two of the seven radiogalaxies,while Hβand Hαareﬁlled up by conspicuous emission lines.A clear exception to the presence of the Balmer series in absorption is the broad-line radiogalaxy3C382.In order to estimate the Balmer strength,crucial to date any young stellar population involved,we use the EW of the H10λ3798˚A line,which appears only weakly contaminated by emission in the radiogalaxies.H10is chosen as a compro-mise of an easily detectable Balmer line that shows both a minimum of emission contamination and clear wings to mea-sure the adjacent continuum.The Balmer lines from Hβto H9λ3836˚A are contaminated by prominent emission,which in Case B recombination comes in decreasing emission ra-tios to Hβof1,0.458,0.251,0.154,0.102,0.0709(Osterbrock 1989);H10has an emission contamination of0.0515×Hβ. At the same time,the absorption strengths are quite similar from Hβto H10,although the EW(H10)is actually sys-tematically smaller than EW(Hβ)in young to intermediate-age populations.Gonz´a lez-Delgado,Leitherer&Heckman (2000)obtain,in their population synthesis models,ratios of EW(Hβ)/EW(H10)between1.3and1.6for bursts with ages0to1Gyr and constant or coeval star formation histo-ries.Lines of order higher than10have decreasing emission contamination,but they also increasingly merge towards the Balmer continuum limit.A caveat in the use of H10as an age calibrator comes from the fact that this line might be contaminated by metal-lic lines in old populations.Although our measurements of H10in NGC4374are around1.5˚A,an inspection of the spec-tra of three elliptical galaxies(NGC584,NC720,NGC821) observed in the same wavelegth range(but with lower S/N) and archived in the Isaac Newton Group database,indicates that a wide range of EW(H10),from2to4˚A,could char-acterize elliptical galaxies,while their Hβindices are in the 1to2˚A regime.If conﬁrmed by better data,these results could indicate that although the upper Balmer series is de-tected in elliptical galaxies,it could indeed be contaminated by the absorptions of other species.Clearly,more work needs to be done in the near-UV spectra of elliptical galaxies be-fore conclusive evidence can be derived for the behaviour of EW(H10)in old stellar systems,and its contamination by metallic lines.In all the radio-galaxies observed in this work,the H10 proﬁle is narrow and reproduces the shape of the wings of the lower-order Balmer absorption lines.Hydra A and3C285 clearly provide the bestﬁttings.As an illustration,Figure6 shows the estimated absorption line proﬁles for the Hβ,Hγand Hδlines,assuming a constant ratio between their EWs and that of H10,and also a scaled(×1.4)H10proﬁle for the case of Hβin Hydra A.We also measured indices that are mostly sensitive to the metal content of the stellar populations involved. The Lick indices of Mg and Fe(e.g.Worthey et al.1994) serve this purpose.In order to avoid the contribution of [O III]λ4959˚A to the continuum measurement for the molec-ular index Mg2,we have displaced the lower continuum band of this index to4895.125˚A<λ<4935.625˚A.This redeﬁni-tion does not alter the value of the index in the elliptical galaxy,which shows no[O III]emission.Table4compiles the EW of H10,and the metallic in-dices Mg b,Fe5270,Fe5335,[MgFe],Mg2of the Lick system, measured in the rest-frame of the galaxies in our sample.The atomic indices are aﬀected by broadening,like the CaT in-dex,while Mg2is only aﬀected by lamp contributions in the original IDS Lick system(Worthey et al.1994,Longhetti et al.1998).We have calculated broadening corrections as in section3.1for the atomic lines,and adopted the correc-tions calculated by Longhetti(1998)for the molecular lines. The uncorrected values of these indices are denoted with a subindex u in Table4.The errors of the individual line and molecular indices were estimated adopting continua shifted from the bestﬁt continua by±1σ.This lead to average er-rors between an8and a10per cent for individual line and molecular indices,and∼6per cent for[MgFe].The agreement between our measurements of Lick in-dices and those carried out by other authors(Gonz´a lez1993, Davies et al.1987,Trager et al.2000a)on our galaxy in com-mon,NGC4374,is better than10per cent.4.DISCUSSIONc 0000RAS,MNRAS000,000–000Stellar populations in the nuclear regions of nearby radiogalaxies5parison with elliptical galaxies andpopulation synthesis modelsThe analysis of the spectral energy distributions and colours of elliptical galaxies suﬀers from a well known age-metallicity degeneracy(Aaronson et al.1978).However,this is broken down when the strengths of suitable stellar absorption lines are taken into account(e.g.Heckman1980).The plane com-posed by the[Hβ]and[MgFe]indices,in this sense,can dis-criminate the ages and metallicities of stellar systems.It is on the basis of this plot,that a large spread of ages in ellip-tical galaxies has been suggested(Gonz´a lez1993).Bressan, Chiosi&Tantalo(1996)claim,however,that when the UV emission and velocity dispersion of the galaxies are taken into account,the data are only compatible with basically old systems that have experienced diﬀerent star formation histories(see also Trager et al.2000a,2000b).A recent burst of star formation that involves only a tiny fraction of the whole elliptical mass in stars,would rise the[Hβ]index to values characteristic of single stellar populations which are 1to2Gyr old(Bressan et al.1996).Most likely,the stellar populations of radiogalaxies are also the combination of diﬀerent generations.Direct sup-port for this interpretation in the case of Hydra A comes from the fact that the stellar populations responsible for the strong Balmer lines are dynamically decoupled from those responsible for the metallic lines(Melnick,Gopal-Krishna& Terlevich1997).This interpretation is also consistent with the modest ∆4000˚A measurements we have obtained.Figure7shows a comparison of the values found in radiogalaxies,with those of normal elliptical,spiral and irregular galaxies,including starbursts,from the atlas of Kennicutt(1992).The radio-galaxies3C98,3C192,4C73.08and DA240have indices of the order of1.9to2.3,which overlap with those of normal E galaxies,∆4000˚A=2.08±0.23.These values correspond to populations dominated by stars of ages1to10Gyr old, if one assumes the coeval population synthesis models of Longhetti et al.(1999).However,Hydra A and3C285have indices in the range1.4to1.6,typical of coeval populations which are200to500Myr old.Once the bulge population is subtracted,the∆4000˚A indices of Hydra A and3C285 decrease to1.2and1.0respectively,which are typical of systems younger than about60Myr.Hamilton(1985)measured the∆4000˚A index in a sam-ple of stars covering a wide range of spectral types and lu-minosity classes.He found a sequence of increasing∆4000˚A from B0to M5stars,with values from1to4mag respec-tively.A comparison with the sequence he found leads us to conclude that the break in the bulge subtracted spectrum of Hydra A is dominated by B or earlier type stars while that of3C285is dominated by A type stars.The index ∆4000˚A does not clearly discriminate luminosity classes for stars with spectral types earlier than G0.The equivalent width of the H10absorption line in these two radiogalaxies give further support to the inter-pretation of the Balmer break as produced by a young stel-lar population.In Hydra A weﬁnd after bulge subtraction EW(H10)≈3.9˚A,which,according to the synthesis models of Gonz´a lez-Delgado et al.(2000)gives ages of7to15Myr for an instantaneous burst of star formation,and40to 60Myr for a continuous star formation mode,in solar metal-licity environments.In the case of3C285,EW(H10)≈6˚A would imply an age older than about25Myr for a single-population burst of solar metallicity.The metallic indices of normal elliptical galaxies range between the values0.56<∼log[MgFe]<∼0.66(Gonz´a lez 1993),which characterizes oversolar metallicites for ages larger than about5Gyr.This is also the typical range of our radiogalaxies,although3C285shows a clear departure with log[MgFe]≈0.4.However,[MgFe]tends to be smaller for populations younger than a few Gyr and similar overso-lar metallic content.Since3C285has a clear burst of recent star-formation,we conclude that its overall abundance is also most probably solar or oversolar.4.2.The blue stellar contentA better estimate of the spectral type and luminosity class of the stars that dominate the break in Hydra A and3C285 comes from the two-dimensional classiﬁcation of Barbier and Chalonge.In Figure8the solid squares connected by lines represent the maximum range of possible Dλ1values mea-sured in these radiogalaxies.The Balmer break index is sensitive to the position-ing of the pseudo-continuum on top of the higher order Balmer series lines,which in turn is sensitive to the merging of the wings of the lines,enhanced at large velocity dis-persions.In order to assign spectral types and luminosity classes to the stars that dominate the break,therefore,it is not suﬃcient to compare the values we have obtained with those measured in stellar catalogues.The values mea-sured for the radiogalaxies can be corrected for their in-trinsic velocity dispersions;we have chosen instead to recal-ibrate the index using template stars of diﬀerent spectral types and luminosity classes convolved with gaussian func-tions,until they reproduce the width of the Balmer lines observed in the radiogalaxies(FWHM≈12.5˚A).We used the B0to A7stars from the stellar library of Jacoby et al. (1984),which were observed with4.5˚A resolution.The val-ues of the Dλ1indices measured in these broadened stars are represented in Figure8by their respective classiﬁcation. By comparison we also plot the grid traced by the locus of unbroadened stars,as published by Str¨o mgren(1963).The broadening of the lines shifts the original locus of supergiant stars from theλ1<∼3720˚A range(Chalonge1956)to the 3720<∼λ1<∼3740˚A range,occupied by giant stars in the original(unbroadened)classiﬁcation.Giant stars,in turn, shift to positionsﬁrst occupied by dwarfs.Most dwarfs have Balmer line widths comparable to those of the radiogalaxies, and thus their locus in the diagram is mostly unchanged.The value of the D index indicates that the recent burst in Hydra A is dominated by B3to B5stars,and the eﬀective position of the Balmer break(λ1)indicates that these are gi-ant or supergiant stars,respectively.These stars have masses of7and20M⊙(Schmidt-Kaler1982).From the stellar evo-lutionary tracks of massive stars with standard mass-loss rate at Z⊙or2Z⊙(Schaller et al.1992,Schaerer et al.1993, Meynet et al.1994)we infer that these stars must have ages between7to8Myr(B3I)and40Myr(B5III).Note that the B4V stars in Figure8,near the location of Hydra A, cannot originate the break and at the same time follow the kinematics of the nucleus(see section5.3).Any dwarf star located in the stellar disk of Hydra A would show absorptionc 0000RAS,MNRAS000,000–000。

厦门“PEP”2024年11版小学5年级下册第十五次英语第2单元全练全测(含答案)

厦门“PEP”2024年11版小学5年级下册英语第2单元全练全测(含答案)考试时间：100分钟（总分:120）A卷考试人：_________题号一二三四五总分得分一、综合题(共计100题)1、填空题：I enjoy going for ______ (散步) in the evening. It helps clear my mind and relax.2、What is 6 + 2?A. 7B. 8C. 9D. 10答案:B3、What food do we get from cows?A. EggsB. MilkC. CheeseD. All of the above4、听力题：The flowers are ______ (beautiful) in the garden.5、填空题：The bee buzzes around the _______ (花).6、听力题：Star formation occurs in regions of space called molecular _______.7、How many days are there in a week?A. 5B. 6C. 7D. 8A base feels slippery and can turn __________ paper blue.9、听力题：She _____ (likes) to play soccer.10、填空题：My favorite way to celebrate is ______.11、听力题：A _______ is a combination of two or more substances that are not chemically combined.12、In what type of habitat do polar bears live?A. DesertB. ForestC. ArcticD. Savannah13、填空题：The __________ is a famous natural wonder in South America. (亚马逊雨林)14、填空题：The __________ is known for its unique rock formations. (约塞米蒂)15、填空题：A wildcat is a small ________________ (猫).16、How many days are in a week?A. FiveB. SixC. SevenD. Eight17、选择题：What is the name of the famous mountain in Oceania?A. Mount CookB. Mount KosciuszkoC. Mount TaranakiD. All of the above18、填空题：I want to create a ________ with my ideas.19、听力题：The chemical symbol for californium is _____.The owl has excellent _______ (夜间视力).21、听力题：The __________ is a region known for its rich biodiversity.22、听力题：The chemical formula for potassium chloride is ______.23、听力题：The chemical symbol for cobalt is _______.24、Which bird is known for its colorful feathers and ability to mimic sounds?A. CrowB. ParrotC. SparrowD. Eagle答案:B25、听力题：The _____ of a substance refers to its ability to conduct heat.26、听力题：Reptiles lay ______.27、听力题：I want to be a ______ (teacher) when I grow up.28、听力题：A chemical reaction can produce heat, light, or ______.29、填空题：I have a huge collection of _____ (玩具).30、听力题：The ______ is a common pet that barks.31、听力题：A covalent bond is formed when atoms __________ electrons.32、What do you call the place where you live?A. HouseB. HomeC. CityD. Street答案: BThe girl loves to ________.34、填空题：My favorite ________ is yellow.35、填空题：When light passes through a prism, it creates a _______. (光谱)36、What is the smallest continent?A. AfricaB. AsiaC. AustraliaD. Europe答案:C37、填空题：My brother is a _____ (学生) who loves history.38、填空题：A lizard can regenerate its ______ (尾巴).39、选择题：What color are school buses?A. BlueB. GreenC. YellowD. Red40、听力题：A solution is a homogeneous mixture of two or more ______.41、选择题：What is the color of the sun?A. BlueB. YellowC. RedD. Green42、听力题：The flowers are ___. (growing)43、What do you call a young duck?A. DucklingB. ChickC. GoslingD. Fawn44、听力题：The sun is ______ (shiny) during the day.45、听力题：Granite contains minerals like quartz, feldspar, and ______.46、oasis) is a fertile area in a desert. 填空题：The ____47、听力题：The __________ is a small area that is surrounded by a larger area.48、填空题：I like to __________ (动词) my __________ (玩具名) at night.49、听力题：The chemical formula for acetone is ______.50、填空题：The musician brings joy through _____ (音乐).51、听力题：I have a ___ (dream) of being an astronaut.52、填空题：I enjoy _______ (参加)社会活动。

The stellar content of the Local Group dwarf galaxy Phoenix

a rXiv:as tr o-ph/99539v125May1999The stellar content of the Local Group dwarf galaxy Phoenix 1D.Mart´ınez-Delgado Instituto de Astrof´ısica de Canarias,E-38200La Laguna,Canary Islands,Spain e-mail:ddelgado@ll.iac.es C.Gallart Observatories of the Carnegie Institution of Washington,813Santa Barbara St.,Pasadena,CA 91101,USA e-mail:carme@ and A.Aparicio Instituto de Astrof´ısica de Canarias,E-38200La Laguna,Canary Islands,Spain e-mail:aaj@ll.iac.es ReceivedABSTRACTWe present new deep V I ground-based photometry of the Local Group dwarf galaxy Phoenix.Our results conﬁrm that this galaxy is mainly dominated by red stars,with some blue plume stars indicating recent(100Myr old)star formation in the central part of the galaxy.We have performed an analysis of the structural parameters of Phoenix based on an ESO/SRC scanned plate,in order to search for diﬀerentiated components. The elliptical isopleths show a sharp rotation of≃90deg of their major axis at radius r≃115′′from the center,suggesting the existence of two components: an inner component facing in the east–west direction,which contains all the young stars,and an outer component oriented north–south,which seems to be predominantly populated by old stars.These results were then used to obtain the color–magnitude diagrams for three diﬀerent regions of Phoenix in order to study the variation of the properties of its stellar population.The young population located in the central component of Phoenix shows a clear asymmetry in its distribution,with the younger blue plume stars predominantly located in the western half of the central component and the older core helium-burning stars predominantly situated in the east.This spatial variation could indicate a propagation of star formation across the central component.The H i cloud found at∼6′Southwest by Young&Lo(1997) could have been involved in this process.We alsoﬁnd a decreasing gradient in the density of intermediate-age population with the galactocentric radius, based on the number of stars populating the red clump in the color–magnitude diagram.Since no metallicity gradient is apparent,this indicates the presence of a substantial intermediate-age population in the central region of Phoenix that would be less abundant or absent in its outer regions.This result is alsoconsistent with the gradient found in the number of horizontal branch stars, whose frequency relative to red giant branch stars increases towards the outer part of the galaxy.These results,together with those of our morphological study,suggest the existence of an old,metal-poor population with a spheroidal distribution surrounding the younger inner component of Phoenix.Thistwo-component structure may resemble the halo–disk structure observed in spirals,although more data,in particular on kinematics,are necessary to conﬁrm this.We have estimated the average star formation rate for the last1Gyr and for the age interval1–15Gyr from the number of blue and red giant branch and asymptotic giant branch stars observed in the color–magnitude diagram.For the central region,the average past star formation rate is very similar to that for the last1Gyr.The recent star formation rate of Phoenix is also comparable to that displayed by typical dIrr galaxies except perhaps for the fact that it lacks any strong very recent burst as exhibited by galaxies such as Sextans A or NGC6822.The area-normalized star formation rate for the central region of Phoenix is in the range obtained by Hunter&Gallagher(1986)for their sample of dIrr galaxies.We have determined a distance modulus for Phoenix of(m−M)0=23.0±0.1 using the tip of the red-giant branch as a distance indicator.Weﬁnd four short-period variable candidates from our photometry that might be anomalous Cepheids or W Vir stars.Finally,it is very unlikely that Phoenix has globular clusters,as is expected for a galaxy with such a faint absolute magnitude.Subject headings:galaxies:dwarf—galaxies:Local Group—galaxies:stellar content—galaxies:structure—techniques:photometry1.IntroductionIn generic hierarchical clustering scenarios for galaxy formation,such as cold dark matter dominated cosmologies(White&Rees1978;Blumenthal et al.1984;Dekel&Silk 1986),dwarf galaxies should have formed prior to the epoch of giant galaxy formation and would be the building blocks of larger galaxies.The dwarf galaxies observed today would be surviving objects that have not merged with larger galaxies,and their underlying structure could provide important clues about the process of dwarf galaxy formation at high redshifts.Local Group dwarf galaxies oﬀer the only opportunity to study their evolution in great detail through the observation of their resolved stellar populations,which provide fossil records of their star formation history(SFH)and chemical enrichment.In addition, detailed observations of their two-dimensional structure,designed to search for gradients in the stellar populations,permit the determination of the internal structure of these systems and provide direct input for detailed models of dwarf galaxy formation.In this paper,we present observations of the Phoenix dwarf galaxy,providing a full spatial coverage of the system that allows us to study the stellar population gradients of the galaxy.The Phoenix dwarf galaxy was discovered by Schuster&West(1976)from visual inspection of a B plate of the ESO survey.Theﬁrst photographic photometry by Canterna &Flowers(1977),which showed the presence of some young blue stars,suggested this low surface brightness object was a distant dwarf irregular galaxy.Deeper color–magnitude diagrams(CMDs)obtained by Ortolani&Gratton(1988)brought Phoenix into the Local Group,at a distance of about500kpc,and showed that it is a dwarf galaxy dominated by an old,metal-poor population,with a small number of young stars.These authors proposed that Phoenix belongs to an intermediate class between typical irregular galaxies and dwarf spheroidals,because it shares properties of both types of galaxies.Similar conclusions were reached by Van de Rydt,Demers&Kunkel(1991,VDK hereafter)in their BV Iphotometry study of the stellar content of Phoenix.They found Phoenix to be still closer (∼400kpc)to the Milky Way and argued that the few blue stars observed in its CMD can be explained as a burst of star formation produced less than150Myr ago.The possible presence of upper asymptotic giant branch(AGB)stars also suggested the existence of an intermediate-age population and hence a complex star formation history for this galaxy.Phoenix is situated on the west edge of the Magellanic Stream,whose H i structure is complex and multivalued near the position of Phoenix(Morras1985).Therefore,although several H i surveys have detected small amounts of gas near Phoenix,the current lack of an optical velocity for the galaxy means that an association between the H i and Phoenix is speculative.Carignan et al.(1991)detected H i emission at V⊙=56km s−1,clearly separated from a much larger scale component at∼120km s−1associated with the Magellanic Stream previously detected by Morras&Bajaja(1986).Oosterloo,Da Costa &Staveley-Smith(1996)reported that the H i detected by Carignan et al.(1991)is part of a large H i complex extending over more than one degree and is thus not likely to be associated with Phoenix.These authors detected a smaller cloud at V⊙=−35km s−1, situated at6′SW from the center of the galaxy,which could be associated with Phoenix. Recent VLA observations by Young&Lo(1997)ahve conﬁrmed the existence of this cloud (cloud A)close to the optical body of the galaxy,which forms a curved structure that wraps around its SW border.Furthermore,they conﬁrmed there is no H i coincident with the optical body and detected additional H i about10′south of it,with V⊙=55km s−1.Hence, it is very important to measure the optical radial velocity of Phoenix in order to conﬁrm any possible association with these H i clouds.It that were the case,it would be possible to obtain the M HI/L B relation,which is an important parameter necessary for understanding the evolutionary status of Phoenix.In this paper,we discuss the stellar content of Phoenix in the light of new ground-basedV I CCD photometry.Brief progress reports of this study were given by Mart´ınez-Delgado, Gallart&Aparicio(1998),Mart´ınez-Delgado&Aparicio(1998a)and Mart´ınez-Delgado et al.(1999).In Sec.2we present the observations,data reduction,proﬁle-ﬁtting photometry and artiﬁcial-star tests.Sec.3discusses the spatial structure of Phoenix on the basisof digitized photographic material.In Sec.4we present the CMD of Phoenix from our CCD photometry.In Sec.5we calculate its distance using the tip of the red giant branch (TRGB).Sec.6investigates the spatial variation of the diﬀerent stellar components of Phoenix.In Sec.7we study the variable stars.In Sec.8we discuss the possible existence of globular clusters in Phoenix.Finally,in Sec.9the results presented throughout the paper are summarized.2.Observations and data reductionObservations of the Phoenix dwarf galaxy were obtained in1997February at the du Pont100′′telescope at Las Campanas Observatory(LCO)under photometric conditions. The detector used was a2048×2048thinned Tektronix chip,which provided a pixel size of0.296′′and a totalﬁeld of8′.Theﬁeld was centered4′north west from the center of Phoenix with the purpose of including the external regions of the galaxy and avoiding the bright star situated close to its border.Images in V and I bands were obtained during nine consecutive nights as part of a strategy to search for Cepheid variable stars.The best seeing images(<1.0′′)from this campaign were selected for this study.The total integration times were2100s in V and 2700s in I.In addition,1200-s exposures in both V and I of a comparisonﬁeld situated 7.5′east of the center of Phoenix were taken with the purpose of studying the foreground contamination.The journal of observations for the frames selected for this study is given in Table1.A B image was taken during a diﬀerent observing run at LCO to obtain the colorimage of the central region of Phoenix shown in Figure1from the B,V and I images using the IRAF task RGBSUN.The IRAF CCDRED package was used to correct the data for overscan andﬂatﬁeld, using bias and high signal-to-noise skyﬂats taken each night.The I images were corrected from the usual interference pattern using a fringe frame obtained from the median of a set of nearly blankﬁeld images acquired during the observing run.This fringe image was then suitably scaled and removed from the I images.2.1.PhotometryThe photometry of the stars in Phoenix has been obtained using the ALLFRAME software(Stetson1994),kindly made available to us by Dr.Stetson.A medianed frame was obtained from the combination of the seven V and seven I images selected for this study.The DAOPHOT II/ALLSTAR package was used to obtain a complete list of stars on the medianed image,using three FIND/ALLSTAR/SUBSTAR passes.This list was given as input to ALLFRAME to perform the photometry in14individual frames.Only stars measured in at least three images were retained.PSFs computed in each individual frame were used in this process.We adopted a Moﬀat function withβ=2.5plus a table of residuals which vary quadratically with position in the frame.DAOMASTER was used to obtain the instrumental V and I magnitudes as the error-weighted average of the magnitudes for each star in the individual frames,selecting only stars withσV<0.2andσI<0.2.Next,we paired the V and I list to obtain(V−I) color indexes.Theﬁnal number of measured stars is6530.Theσstandard errors of the mean of these stars are plotted in Figure2.Atmospheric extinction corrections for each night and the photometric transformationto the Johnson–Cousins standard system were derived from observations during the campaign of a total of156measures in each band V and I of37standard stars from the list of Landolt(1992).Theﬁnal equations used to transform our instrumental magnitudes into the Johnson–Cousins magnitudes are(V−v)=24.517−0.027(V−I)(I−i)=24.340−0.029(V−I)The photometric conditions during the campaign were stable and produce very small zero point errors for the photometric transformation:±0.006mag in V and±0.008mag in I.The standard errors in the extinction were always smaller than0.012in both V and I. Aperture corrections were obtained from a large number of isolated stars in one V and one I image of Phoenix.These are accurate to±0.010mag in V and±0.009mag in I.Taking all these uncertainties into account,we estimate a total error in the photometry zero point of0.02mag in both V and I.2.2.Artiﬁcial Stars TestsArtiﬁcial star tests have been performed in a way similar to that described in Aparicio &Gallart(1995)and Gallart et al.(1996a)to obtain the necessary information about the observational eﬀects present in the photometry.In short,a number of artiﬁcial stars of known magnitudes and colors are injected in the original frames(to produce what we call synthetic frames)using the ADDSTAR algorithm of DAOPHOT II(Stetson1993),and the photometry is done again following the same procedure used to obtain the photometry for the original frames.This process is repeated as many times as necessary to test alarge number of artiﬁcial stars(25000in this particular case).The injected and recovered magnitudes of the artiﬁcial stars,together with the information from those that were lost, provide the necessary information for all the observational eﬀects present in the photometry (see Aparicio&Gallart1995for a thorough discussion of these eﬀects).The distribution of the stars in the CMD has been chosen in a way similar to that of Mart´ınez-Delgado&Aparicio(1998b),using a synthetic CMD constructed with constant star formation rate(SFR;from15to0.01Gyr)and metallicity,Z,linearly increasing with time from0to0.002.The main diﬀerence from the former paper is that the procedure here has been adapted to the current data reduction performed with ALLFRAME,and is a longer and more complicated procedure.Because each artiﬁcial star trial requires a large amount of CPU time,we have optimized the distribution of artiﬁcial stars in the real frames by compromising between adding the largest possible number of stars to each frame and retaining its original crowding characteristics.All we require for the artiﬁcial-star tests to represent the actual crowding and other observational eﬀects of our images is that the artiﬁcial stars do not interact with each other,i.e.,that the recovered magnitudes are aﬀected only by the stars in the galaxy plus any other artifacts in the image possibly aﬀecting the photometry.We constructed a grid with(x,y)coordinates of the artiﬁcial stars equally spaced within it.This grid is displaced randomly for each test to assure a sampling of the observational eﬀects over the frame.The distance between the nodes of the grid is such that the eﬀect on each artiﬁcial star produced by the wings of its artiﬁcial neighbor is negligible.A distance of(2×PSF radius+1)fulﬁls this requirement,because the wings of the artiﬁcial stars extend as far as the PSF radius.With a PSF radius of12pixels in our frames,the artiﬁcial stars have been distributed25pixels apart.With this spacing,we can add approximately6300artiﬁcial stars per trial to each frame,and the number of runs ofartiﬁcial-star tests reduces to four.Figure3shows the injected magnitudes and the resulting recovered magnitudes for the artiﬁcial stars used in this analysis of the observational eﬀects in Phoenix.3.The spatial structure of PhoenixThe study of the structural parameters of dwarf galaxies can provide valuable information about their formation and evolution,as well as the possible relationships between diﬀerent dwarf galaxy types.An open debate exists concerning the possible secular evolution of dIrr to dEs based on the morphological similarities between both types of galaxies,such as the lack of a conspicuous nucleus or the fact that their radial proﬁles can beﬁtted by an exponential law in both cases(Lin&Faber1983;Ferguson&Binggeli 1994).Phoenix is a unique target for approaching this issue because it seems to be a nearby prototype of dIrr/dE.VDK obtained some details about the structure of Phoenix from a scanned ESO/SRC IIIaJ plate of the galaxy.They found that Phoenix follows an exponential surface brightness proﬁle,which reaches the zero level at∼8.7′along the major axis but does notﬁt a King model.They concluded that Phoenix is a rather isolated galaxy which has never been tidally aﬀected by the Milky Way.Apart from this,there is no other detailed study of the structural parameters of Phoenix in the literature.We have performed an analysis of the structural parameters of Phoenix using a method similar to that described by Irwin&Hatzidimitriou(1995,IH hereafter).Our results are presented in this section.3.1.Photographic plate material and star listIt is not possible to derive the global structural parameters of Phoenix from our CCD image because itsﬁeld is not large enough to cover the whole galaxy.We used a portion of a plate from the UK Schmidt Telescope(UKST)survey that Dr.Irwin kindly digitized for us.The exposure time for this plate was60minutes in IIIaJ emulsion with a GG395ﬁlter,reaching a limiting magnitude of B∼22.Figure4shows the scanned region of this plate used in this study.This region covers a35′×35′ﬁeld centered on Phoenix and was digitized at a pixel sampling of1′′using the Automatic Plate Measuring System(APM, Kibblewhite et al.1984)at Cambridge.Because of we are using a central portion of a larger plate(6◦×6◦),the resulting image is quite uniform and there are no systematic magnitude diﬀerences caused by optical vignetting.DAOPHOT II/ALLSTAR software(Stetson1993)was used to make a catalog of the objects detected in the plate.A preliminary list of stars was obtained after choosing an adequate FIND threshold.This list was given as input to ALLSTAR,which performed a proﬁle-ﬁtting photometry using the PSF of the plate.A total of11032stars were measured.3.2.Isopleth maps of PhoenixIsopleth maps of the distribution of stars in a resolved galaxy is a means of analyzing its two-dimensional structure.The density maps of this study have been obtained by means of a kernel method similar to that discussed by Alfaro,Delgado&Cabrera-Ca˜n o(1992).In short,a map with the position of the stars was convolved with a two-dimensional Cauchy kernel with a smoothing parameter of h x=h y=25′′.The result of this process is a smooth map of the spatial stellar density of the galaxy,which gives a visual description of its morphology.However,the density maps derived from our star list are aﬀected by twofactors:i)contamination by foreground Galactic stars and background galaxies,and ii) crowding eﬀects in the inner part of the galaxy.Both eﬀects are diﬃcult to estimate,and in the case of background contamination are unreliable at fainter magnitudes.For this reason we adopted an approach similar to that of IH,using all the raw stellar images detected to construct the isopleth map and correcting a posteriori the obtained radial proﬁle for crowding eﬀects and foreground contamination(see Sec.3.3).Figure5shows the isopleth map obtained for Phoenix.The structural parameters can be determined byﬁtting ellipses to each isodensity contour.This provides a reliable means of deriving the center,ellipticity(ǫ=1−b/a)and position angle(PA)of the number distribution of stars.The parameters obtained by this procedure do not exhibit signiﬁcant variations through the galaxy,except at r∼115′′from the center,where a sharp rotation of∼90◦in the orientation of the semi-major axes of the ellipses is observed. This change suggests the existence of two components in Phoenix:an inner component oriented east–west;and an outer component which is more extended and clearly aligned north–south.Interestingly,the inner component contains the majority of the galaxy’s young stars,with aﬂattened distributionﬁtting its shape quite well(see Sec.6.1).This, together with the clearly diﬀerent orientation from the outer component could indicate that the galaxy has a disk–halo structure similar to that of bigger,spiral galaxies.However, data on the kinematics are necessary before deﬁnitely stating that this is the case.Table2 lists the mean of the center(in pixels),ellipticity and PAs for both components.The outer component was previously detected by VDK in their study of the morphology of Phoenix.They found a similar ellipticity(ǫ∼0.3)and orientation as that given here for this component.3.3.Radial proﬁle and modelﬁttingThe radial density proﬁle of Phoenix can be derived from the original star list described in Sec.3.1by counting stars within suitably oriented elliptical annuli centered on the galaxy center.We will adopt the radial proﬁle obtained using the values of the center,ellipticity and PA for the outer component(Table2)because it allows a more accurate study of the outer regions of Phoenix(see Sec.3.4).As mentioned in Sec.3.2,this raw radial proﬁle must be corrected for observational eﬀects and background contamination(which includes Galactic stars and background galaxies).Evaluation of the background contamination is critical to determining the outer regions of the galaxy and its actual extension(i.e.,its tidal radius).In our case,the contamination by background galaxies is strong due to the presence of a galaxy clusterin theﬁeld of the scanned region.To achieve this correction,we adopted the median of the value of the number density at large radii(r∼23′)as the best estimate of the true background level(d bck).This value,d bck=8.0±0.4(′)−2was subtracted from the raw radial proﬁle of Phoenix.Crowding corrections were obtained from artiﬁcial-star tests in the digitized plate,in a similar way as for the CCD images(see Sec.2.2).A total number of120000stars covering the complete range of instrumental magnitudes of the plate were added to the image in small groups of1200stars.The resulting frames were processed to obtain the magnitudes of the artiﬁcial stars aﬀected by the observational eﬀects,and build a crowding-trial table (see Aparicio&Gallart1995)for the plate.This table was used to derive the crowding factor,Λ,in each elliptical annulus.The crowding factor is deﬁned asΛ=N rc/N in,where N rc is the number of recovered stars and N in is the number of injected stars in each annulus. TheseΛvalues were then used to correct the star counts in each elliptical annulus in order to obtain the crowding-corrected radial proﬁle of Phoenix plotted in Figure5.Fitting models to the radial proﬁles of dwarf galaxies like Phoenix can provide information about whether they have always been isolated or whether,on the contrary,they have been tidally truncated in a past encounter with the Milky Way.Two basic models have been extensively applied to study the radial proﬁles of dwarf galaxies:a)the King model (King 1962),which assumes that the system is isothermal with a Maxwellian distribution of velocities truncated by the presence of a dominant external mass (in our case,the Milky Way);and b)the exponential law,which is based on empirical evidence that exponential proﬁles it galactic disks and Irr galaxies very well (Faber &Lin,1983).In the case of Phoenix,the existence of the inner and outer structural components probably makes it unrealistic to try ﬁtting a model proﬁle at all distances.Furthermore,since the tidal truncation eﬀects,if present,should be much more evident in the outer component (a >112′′)a King ﬁtting has been tried for it.We adopt the simple method of King (1962)rather than determine the tidal and core radii simultaneously.For the outer part,the stellar surface density can be expressed asd (r )=d 1(1/r −1/r t )2,where d 1is a constant and r t is the tidal radius.This relation diﬀers only slightly from the exact King model formula given by King (1962;formula 14)in the outer regions of the galaxy.The value of r t can be obtained from the slope and zero point of a least-square linear ﬁt of a plot of d 1/2against 1/r ,as is shown in Figure 7.This yields r t =15′.8+4′.3−2′.8,where the errors have been estimated after varying d bck by ±1σ.This is the main source of error in the determination of r t and means that the angular size (tidal diameter)of Phoenix is slightly larger than that obtained in the previous study by VDK (r t =8.7′).This result is compatible with our CCD observations,where an evident galaxy stellar population isdetected at a large galactocentric radius(∼8′).It also shows that the outer component of Phoenix can beﬁtting by a King model and might therefore have been aﬀected by the gravitationalﬁeld of the Milky Way.Knowledge of the radial velocity of the galaxy could help in constraining the orbit of Phoenix,and therefore give clues about possible past encounters with the Milky Way.4.The Color–Magnitude Diagram of PhoenixFigure8shows the CMD of Phoenix.It has features in common with the CMDsof both dE and dIrr galaxies(Fornax:Stetson,Hesser&Smecker-Hane1998;Pegasus: Aparicio et al.1997a;Antlia:Aparicio et al.1997c),and displays a number of noteworthy stellar populations that provideﬁrst insights into the star formation history of this galaxy.The most prominent feature is the red tangle(see Aparicio&Gallart1994),which corresponds to the red giant branch(RGB)and asymptotic giant branch(AGB)loci.This shows,in agreement with previous results,that this galaxy is dominated by red stars (Ortolani&Gratton1988;VDK).The color of the RGB stars indicates a low metallicity ([Fe/H]=–1.36;see Sec.6.4.2)and its dispersion may indicate the existence of an abundance range in Phoenix,although it could also be accounted for by dispersion in age only(see Sec.6.4.2).A number of red stars are observed above the tip of the RGB(TRGB),that could be intermediate-age AGB stars.However,the CMD of the comparisonﬁeld shows a large number of foreground stars in this region of the diagram.The existence of this upper-AGB population is discussed in detail in Sec.6.3.1.A blue-plume(BP)composed of young main sequence(MS)stars and/or blue loops is observed at21mag<I<24mag and−0.5<(V−I)<+0.2,indicating that recent star formation has taken place in Phoenix.In addition to this young population,a signiﬁcantnumber of stars are observed at I∼22mag and(V−I)∼0.8,close to merging with the RGB locus.These stars are probably core helium-burning intermediate-mass stars with ages ∼0.8Gyr(see Sec.6.1).Some of the stars observed in the strip0.9<(V−I)<1.1and 21mag<I<18mag may correspond to the reddest extreme of the core helium-burning phase of younger stars.One of the most striking features is the presence of a well-populated red clump(RC) at I∼22.5mag,(V−I)∼1.0,and a blue-extended horizontal branch(HB)extending down to I∼23.5mag,(V−I)∼0.2.The RC appears as a widening of the RGB and is the locus of an old to intermediate-age core helium-burning population.The fact that it extends about1magnitude in luminosity and is well-populated indicates that,at least in its central part(see Sec.6.3.2),the intermediate-age stars make a substantial contribution to the stellar population of Phoenix(it can be compared with Leo I:Gallart et al.1999a). The blue-extended HB is similar to those observed in other dEs(Tucana:Seitzer et al. 1998;Fornax:Stetson et al.1998).Because its position is very close to our limiting magnitude,we cannot characterize it properly,and its existence could even be questioned. Nevertheless,its location coincides with that expected for a galaxy at the distance of Phoenix and it appears horizontal in a[V,(V−I)]CMD.Furthermore,it has also been observed in an independent study of this galaxy(Held,Saviane&Momany1999).In the absence of RR Lyrae variable observations,the HB provides theﬁrst evidence of an old(> 10Gyr),low-metallicity stellar population in Phoenix.5.The distance to PhoenixThe distance to Phoenix has been obtained from the absolute I magnitude of the TRGB(I TRGB),which has been proven to be an excellent distance estimator for nearby resolved galaxies(Lee et al.1993).This point marks the core heliumﬂash of low-mass。

RevModPhys.82.331

CONTENTS
I. The Cosmological Standard Model A. Historical outline B. Friedmann models 1. The metric 2. Redshift and expansion 3. The radiation-dominated phase 4. Age, distances, and horizons C. Structures 1. Structure growth 2. The power spectrum 3. Nonlinear evolution II. The Age of the Universe A. Nuclear cosmochronology 1. The age of the Earth 2. The age of the Galaxy B. Stellar ages C. Cooling of white dwarfs D. Summary III. The Hubble Constant A. Hubble constant from Hubble’s law 1. Hubble’s law 2. The distance ladder 3. The HST key project B. Gravitational lensing C. Summary IV. Big-Bang Nucleosynthesis A. The origin and abundance of 4He 1. Elementary considerations 2. The Gamow criterion 3. Elements produced 4. Helium abundance 5. Expected abundances and abundance trends B. Observed element abundances 1. Principles 2. Evolutionary corrections 3. Speciﬁc results 332 332 333 333 333 334 335 335 335 336 337 337 337 337 338 339 340 340 340 340 340 341 342 342 343 343 343 343 344 344 345 345 346 346 346 346

不同星族恒星的Sc和Mn元素丰度分析

不同星族恒星的Sc和Mn元素丰度分析摘要恒星元素丰度是从观测上追踪银河系结构和化学演化的很好探针。

随着天文观测技术和仪器的改进和提高，近年来人们积累了大量的恒星光谱数据，分析得到了恒星中许多元素的丰度，为我们研究不同类型恒星的化学演化提供了支持。

本文我们以F和G 型矮星作为研究对象，分析这些恒星中Sc和Mn这两种元素的丰度趋势。

为了得到更具普遍性的观测结果，我们采用大样本恒星的统计分析，所选择的样本有2个来源：第一个来源是选择了从2000年以来一直到2011年的11个大样本恒星的观测结果，第二个来源是使用中国科学院国家天文台兴隆观测站的2.16米望远镜，观测得到的15颗F和G型矮星的元素丰度。

总共获得了798颗恒星的元素丰度，包括367颗薄盘星，303颗厚盘星和128颗晕星。

论文首先介绍了银河系的旋臂结构、成分结构，综述了对恒星Sc和Mn元素丰度的观测研究和分析结果；第二部分是星族的分类，用纯运动学来划分薄盘星、厚盘星和晕星；第三部分是我们的光谱观测和丰度分析结果，观测并分析了15颗贫金属矮星的元素丰度；第四部分多样本统计分析，包括样本的选取和完备性分析；最后一部分详细讨论了不同星族恒星的Sc和Mn元素的丰度演化趋势。

通过大样本恒星的统计分析，得到了以下重要结果：（1）Sc元素丰度的演化趋势：贫金属星中[Sc/Fe] 总体上是相对超丰，而且随[Fe/H]的增加呈下降趋势，但在整个金属丰度范围内[Sc/Fe]并不是单一斜率下降，[Sc/Fe]随金属丰度[Fe/H]的演化趋势在薄盘星、厚盘星和晕星中是不完全相同的。

在薄盘恒星（[Fe/H] > -0.8）中[Sc/Fe] 随[Fe/H]的增加呈线性下降趋势，这一特性和α元素很相似，因此Sc有时被称为“类α元素”；在厚盘恒星中当[Fe/H]<-0.5 时[Sc/Fe]随[Fe/H]的增加是线性上升的，但当[Fe/H]达到-0.5后继续增加[Sc/Fe]转而变为线性下降的趋势；在晕族恒星中，[Sc/Fe]的弥散变得较大，且[Sc/Fe] 几乎与[Fe/H]的变化无关。

IRAC Mid-Infrared Imaging of the Hubble Deep Field South Star Formation Histories and Stell

a r X i v :a s t r o -p h /0504219v 1 8 A p r 2005Preprint typeset using L A T E X style emulateapj v.6/22/04IRAC MID-INFRARED IMAGING OF THE HUBBLE DEEP FIELD SOUTH:STAR FORMATION HISTORIES ANDSTELLAR MASSES OF RED GALAXIES AT Z >21I VO L ABBÉ2,J IASHENG H UANG 3,M ARIJN F RANX 4,G REGORY R UDNICK 5,P AULINE B ARMBY 3,E MANUELE D ADDI 6,P IETER G.VAN D OKKUM 7,G IOVANNI G.F AZIO 3,N ATASCHA M.F ÖRSTER S CHREIBER 8,A LAN F.M.M OORWOOD 9,H ANS -W ALTER R IX 10,H UUBR ÖTTGERING 4,I GNACIO T RUJILLO 10,P AUL VAN DER W ERF 4(Accepted for publication in the ApJ Letters)ABSTRACTWe present deep 3.6−8µm imaging of the Hubble Deep Field South with IRAC on the Spitzer Space Tele-scope .We study Distant Red Galaxies (DRGs)at z >2selected by J s −K s >2.3and compare them to a sample of Lyman Break Galaxies (LBGs)at z =2−3.The observed UV-to-8µm spectral energy distributions are ﬁt with stellar population models to constrain star formation histories and derive stellar masses.We ﬁnd that 70%of the DRGs are best described by dust-reddened star forming models and 30%are very well ﬁt with old and “dead”ing only the I −K s and K s −4.5µm colors we can effectively separate the two groups.The dead systems are among the most massive at z ∼2.5(mean stellar mass <M ∗>=0.8×1011M ⊙)and likely formed most of their stellar mass at z >5.To a limit of 0.5×1011M ⊙their number density is ∼10×lower than that of local early-type galaxies.Furthermore,we use the IRAC photometry to derive rest-frame near-infrared J ,H ,and K ﬂuxes.The DRGs and LBGs together show a large variation (a factor of 6)in the rest-frame K −band mass-to-light ratios (M /L K ),implying that even a Spitzer 8µm −selected sample would be very different from a mass-selected sample.The average M /L K of the DRGs is about three times higher than that of the LBGs,and DRGs dominate the high-mass end.The M /L K ratios and ages of the two samples appear to correlate with derived stellar mass,with the most massive galaxies being the oldest and having the highest mass-to-light ratios,similar as found in the low-redshift universe.Subject headings:galaxies:evolution —galaxies:high redshift —infrared:galaxies1.INTRODUCTIONGalaxies at z >2exhibit very diverse properties:they range from the blue Lyman Break Galaxies (LBGs)which are bright in the rest-frame ultraviolet (Steidel et al 1996a,b)to the Dis-tant Red Galaxies (DRGs)which are generally faint in the rest-frame UV and have fairly red rest-frame optical colors.The DRGs are selected in the observers’near-infrared (NIR)by the simple criterion J s −K s >2.3(Franx et al.2003,van Dokkum et al.2003).The variety in the galaxy population at z >2is comparable to that seen in the local universe,where colors range from very blue for young starbursting galaxies to very red for old elliptical galaxies.An urgent question is what causes the red colors of DRGs at z >2.Are they “old and dead”,or are they actively star forming and more dusty than U-dropout galaxies?First analyses of the optical-to-NIR Spectral Energy Dis-tributions (SEDs)and spectra have suggested that both ef-1Based on observations with the Spitzer Space Telescope ,which is op-erated by the Jet Propulsion Laboratory,California Institute of Technology under NASA contract 1407.Support for this work was provided by NASA through contract 125790issued by JPL/Caltech.Based on service mode ob-servations collected at the European Southern Observatory,Paranal,Chile (LP Program 164.O-0612).Based on observations with the NASA/ESA Hub-ble Space Telescope ,obtained at the Space Telescope Science Institute which is operated by AURA,Inc.,under NASA contract NAS5-265552Carnegie fellow,OCIW,813Santa Barbara Street,Pasadena,CA 91101[e-mail:ivo@ ]3CFA,60Garden Street Cambridge,MA 021384Leiden Observatory,P.O.Box 9513,NL-2300RA,Leiden,The Nether-lands5Goldberg fellow,NOAO,950N.Cherry Ave,Tucson,AZ 857196Spitzer fellow,NOAO,950N.Cherry Ave,Tucson,AZ 857197Department of Astronomy,Yale University,P.O.Box 208101,New Haven,CT 06520-81018MPE,Giessenbackstrasse,D-85748,Garching,Germany 9ESO,D-85748,Garching,Germany10MPIA,Königstuhl 17,D-69117,Heidelberg,Germanyfects play a role:they have higher ages,contain more dust,and have higher mass-to-light (M /L )ratios in the rest-frame optical than LBGs (Franx et al.2003,van Dokkum et al.2004,Forster Schreiber et al.2004;F04).Furthermore,many have high star formation rates >100M ⊙yr −1(van Dokkum et al.2004,F04).Unfortunately,the number of DRGs with rest-frame optical spectroscopy is very small.Inferences in-evitably depend on SED ﬁtting,which has large uncertainties (see e.g.,Papovich,Dickinson,&Ferguson 2001).The SED constraints on the stellar and dust content are expected to im-prove signiﬁcantly by extending the photometry to the rest-frame near-infrared.Here we present the ﬁrst results on rest-frame NIR pho-tometry of DRGs and LBGs in the Hubble Deep Field South (HDFS)as observed with the Infrared Array Camera (IRAC;Fazio et al.2004)on the Spitzer Space Telescope .Where necessary,we assume an ΩM =0.3,ΩΛ=0.7,cosmology with H 0=70km s −1Mpc −1,and a Salpeter Initial Mass Function (IMF)between 0.1and 100M ⊙.Magnitudes are expressed in the AB photometric system unless explicitly stated otherwise.2.THE OBSERV ATIONS,PHOTOMETRY ,AND SAMPLE SELECTIONWe have observed the 5arcmin 2HDFS/WFPC2ﬁeld with the IRAC camera integrating 1hour each in the mid-infrared (MIR)3.6,4.5,5.8,and 8µm −bands.The observations,data reduction,and photometry will be described by Labbéet al.(in preparation).Brieﬂy,we reduced and calibrated the data using standard procedures (e.g.,Barmby et al.2004).The limiting depths at 3.6,4.5,5.8,and 8µm are 25.0,24.6,22.6,and 22.5mag,respectively (5σ,3′′diameter aperture).We PSF-matched the IRAC images and a very deep K s −image (Labbéet al.2003)to the 8µm −band.Nearby sources were ﬁtted and subtracted to avoid confusion (see Labbéet al.2004).We measured ﬂuxes in 4.′′4diameter apertures.The K s −IRAC colors were then combined with the HDFS-catalog2Labbéet al.of Labbéet al.(2003)resulting in 11-band photometry from 0.3to 8µm.In addition to the photometric errors,we add in quadrature an error of 10%to reﬂect absolute calibration un-certainties of IRAC.We selected DRGs in the ﬁeld of the HDFS using the crite-ria of Franx et al.(2003)yielding 14galaxies with photomet-ric redshifts (Rudnick et al.2003)in the range 1.9<z <3.8.One blended source was excluded.In addition,we selected isolated LBGs in the same ﬁeld to the same K s -band limit and in the same redshift range,using the U-dropout criteria of Madau et al.(1996)on the WFPC2imaging (Casertano et al.2000).The two samples are complementary,with only 1source in common.For both samples,the mean redshift is z =2.6with a standard deviation of 0.511.F IG .1.—Observed I −K s versus K s −4.5µm color-color diagram of two samples of z >2galaxies.Distant Red Galaxies (DRGs;ﬁlled circles and stars ),occupy a different color region than z ∼2.5Lyman Break Galaxies (LBGs;diamonds ).DRGs that are best described by an old single-age burst model (SSP;stars )have colors distinct from those better described by con-stant star-forming models (CSF)and dust,indicating that IRAC ﬂuxes may help in separating these populations.The curves show the color-evolution tracks of Bruzual &Charlot (2003)models at a ﬁxed z =2.6:an SSP model with ages ranging from 0.3to 3Gyr (dashed line )and two CSF models with ages ranging from 0.1to 3Gyr,and reddenings of A V =0.6and A V =2.0respectively (solid lines ).The vector indicates a reddening of A V =1mag for a Calzetti et al.(2000)law.3.RED GALAXIES AT Z >2:OLD OR DUSTY OR BOTH?In Fig.1we analyze the I −K s versus the K s −4.5µm col-ors.The K −4.5µm color has a sufﬁciently large baseline and much higher signal-to-noise ratio than the K s −5.8µm or K s −8µm colors.The DRGs and LBGs lie in very differ-ent regions,with little overlap.The mean K −4.5µm color of the DRGs is signiﬁcantly redder than that of LBGs,con-ﬁrming that DRGs have higher M /L ratios in the rest-frame optical.The red I −K s and K −4.5µm colors of the DRGs imply that they must be prominent in IRAC selected sam-ples.Indeed,the majority of red z −3.6µm sources selected by Yan et al.(2004)in the GOODS survey satisfy the DRG selection criteria.We show color tracks for stellar popula-11The photometric redshift accuracy of DRGs from other studies is |z spec −z phot |/(1+z spec )≈0.1(F04,Wuyts et al.in preparation).The esti-mate is based on direct comparison of 16galaxies with both photometric and spectroscopic redshifts.Uncertainties in rest-frame magnitudes and model parameters are based on Monte Carlo simulations that take into account un-certainties in the observed ﬂuxes and redshift estimates.tion models (Bruzual &Charlot 2003;BC03),redshifted to a ﬁxed z =2.6:a single-age burst model (SSP)and two mod-els with constant star formation (CSF),and reddenings typical for LBGs (A v =0.6,Shapley et al 2001),and typical for DRGs (A V =2.0,F04).The K s −4.5µm colors of LBGs are consis-tent with low reddening models,whereas most of the DRGs lie in the area of models with substantial extinction.3.1.Dead galaxies at z =2−3Three of 13DRGs (indicated by star symbols in Fig.1)lie well outside the area of constant star-forming models.Their I −K s colors are too red for their K −4.5µm color and they lie close to the line of an old single-age burst model with ages of 2−3Gyr.The candidate old,passive galaxies are shown in Fig.2.Their SEDs are very well ﬁt by an old single burst population.The gray model curves are predictions based on ﬁts to the exceptionally deep optical/NIR data only,demon-strating that the Spitzer ﬂuxes lie very close to the expected values.Dead galaxies have blue K −4.5µm colors and with Spitzer we can now effectively separate them from dusty star-forming DRGs,which are red in K −4.5µm.Models with ongoing star-formation and dust-reddening ﬁt the SEDs badly.Star-formation histories (SFHs)with expo-nentially declining star-formation rates (SFRs)and timescales >300Myr can be ruled out at the 99.9%conﬁdence level.For the marginally acceptable models,the ratio of on-going to past-average SFR is SFR (t )/<SFR >=0.001indicating that these 3galaxies are truly “red and dead”.The best-ﬁt ages and implied formation redshifts depend on the assumed metallicity Z ,as expected.Super-solar Z =0.05models give a mean age of <t >=1.4Gyr and mean formation redshift <z f >=5,solar Z =0.02models yield <t >=2.9Gyr and <z f >≫5,while sub-solar metallicity models fail to provide good ﬁts to the IRAC ﬂuxes as they are too blue.Hence we in-fer from the models that the “dead”galaxies had formed most of their stellar mass by z ∼5(Z =0.05)or much earlier z ≫5(Z =0.02).The number density of “dead”galaxies is 1.9×10−4h 370Mpc −3,assuming a top hat redshift distribu-tion between z =2and 3.For stellar masses >0.5×1011M ⊙and the same IMF,that number density is 10×lower than that of early-type galaxies in the nearby universe (Bell et al.2003).Obviously,the sampled volume is too small to allow ﬁrm conclusions,but the result is indicative of strong evolution of passive galaxies from z =2.5to z =0.3.2.Dusty star forming galaxiesIn Fig.2we show the SEDs for 3of the 8DRGs whose optical/NIR ﬂuxes were better ﬁt with CSF models and red-dening by a Calzetti et al.(2000)dust law.The observed MIR ﬂux points are often different from predictions based on shorter wavelengths,especially where the ﬂux density F λwas still rising at the observed K s −band.Hence,for the very dusty DRGs (A V >1.5)the MIR ﬂuxes help to better constrain the age and dust in the models.The average ages and extinction of the ﬁts changed mildly after inclusion of the IRAC data:from <t >=1.1to 1.3Gyr,and <A V >=1.9to 1.5mag,re-spectively.Finally,two remaining galaxies could not be ﬁt well by any model,most likely due to emission line contamination.One of them is spectroscopically conﬁrmed as a very strong Lyman-αemitter (Wuyts et al.in preparation).However,the generally good ﬁts to the UV-to-8µm SEDs of the DRGs in-dicate that the red J s −K s colors were caused by old age andIRAC observations of red z >2galaxies3F IG .2.—The UV-to-8µm spectral energy distributions (SEDs)of Distant Red Galaxies (DRGs).The best-ﬁt Bruzual &Charlot (2003)models to the full SEDs (black solid lines )are shown,but also predictions based on the ultradeep optical and near-infrared ﬂuxes only (gray solid lines ).Top row:the 3DRGs best ﬁt with an single-age burst (SSP)model.The mid-infrared (MIR)IRAC ﬂuxes (ﬁlled circles )directly conﬁrm the predictions.Old and dead DRGs have bluer K s −4.5µm colors than very dusty star-forming DRGs.Galaxy 767shows a ﬂux excess at 8µm,possibly related to AGN activity.Bottom row:the MIR predictions for galaxies better ﬁt with constant star forming models and Calzetti et al.(2000)dust-reddening.For highly-reddened galaxies,the MIR observations can be very different from the predictions.Here IRAC ﬂuxes help to better constrain the total age,dust content,and stellar mass in the models.dust-reddened star light,and not other anomalies.We note that one of the dead galaxies (see Fig.1)has an apparent ﬂux excess at 8µm,suggesting the presence of an obscured AGN which starts to dominate the ﬂux in the rest-frame K −band (see e.g.,Stern et al.2005).4.MASS-TO-LIGHT RATIO V ARIATIONS AT Z =2−3We next study the M /L ratios inferred from the SED ﬁts described above,adopting the best ﬁtting SFH (SSP or CSF).Figure 3a shows the modeled M /L K (rest-frame K )versus rest-frame U −V .The curves show the dust-free BC03model M /L K ratios.Galaxies generally lie to the red of the model curves as a result of dust attenuation.The ﬁts imply a large range in M /L K for DRGs and LBGs together (a factor of 6).Furthermore,the M /L K ratio correlates with rest-frame U −V color,where galaxies red in the rest-frame optical have higher M /L K ratios than blue galaxies.The average M /L K of the DRGs is 0.33±0.04,about three times higher than that of the LBGs.Figure 3b shows the derived M /L K against the derived age.The M /L K correlates well with the age,as expected from the models.The role of extinction is greatly reduced in the rest-frame K −band,implying that the differences in the M /L K ratios for DRGs and LBGs are driven by age differences.Figure 3c shows the M /L K ratio against stellar mass M ∗.DRGs dominate the high-mass end.The highest-mass galax-ies (>0.5×1011M ⊙)in this sample all have high M /L K ra-tios,and here the M /L K ratio does not depend strongly on M ∗.At lower masses galaxies have much lower M /L K ratios.This may be a selection effect caused by our magnitude cut-off,as we would miss low-mass galaxies with high M /L K .How-ever,high-mass galaxies with low M /L K would be detected if they existed,hence the lower envelope of the M /L K versus M ∗distribution is real.Two intriguing possibilities are that the mean M /L K decreases towards lower stellar mass,or that the intrinsic scatter in M /L K increases.Our DRGs and LBGs were selected in the observed K s −band (rest-frame V −band).We ﬁnd a large variation inM /L V (a factor of 25),hence selection effects play a major role in NIR studies at high redshift.Selection using MIR ob-servations with Spitzer would improve the situation,but the wide range in M /L K ratios and ages found here indicates that even a IRAC 8µm −selected sample would still be very differ-ent from a mass-selected sample.5.DISCUSSION AND CONCLUSIONSWe have presented rest-frame UV-to-NIR photometry of a sample of DRGs and LBGs at z =2−3in the HDFS.These galaxies span a wide range in properties,similar to low-redshift galaxies.The rest-frame NIR photometry from IRAC helps signiﬁcantly:ﬁrst,by allowing us to separate “old and dead”from dusty star forming DRGs using only the observed I ,K s ,and 4.5µm-band,and second,by improving model con-straints on the heavily obscured DRGs.The wide range in galaxy properties at z =2.5raises sev-eral important issues.First,it demonstrates it is impossible to obtain mass-selected samples photometrically.Even in the rest-frame K -band,the M /L ratio varies by a factor of 6for the DRGs and LBGs in the HDFS.Second,we need to under-stand what produces these variations.If a relation between to-tal stellar mass and M /L K exists (see Fig.3c)then stellar mass might be driving the variations.Deeper IRAC data is needed to establish this well,as incompleteness effects may play a role.It is tantalizing that Kauffmann et al.(2003)ﬁnd a simi-lar correlation at z =0.1,with the most massive galaxies being the oldest and having the highest M /L ratios.The authors also found that above a stellar mass M ∗=6×1010M ⊙galaxy prop-erties correlate only weakly with M ∗,similar to what we ﬁnd at z =2−3.The simplest explanation is that we observe the same galaxies at z =0.1and z =2.5,although we note that hi-erarchical formation scenarios predict signiﬁcant merging for galaxies at z >2.Alternatively,we observe merging processes occurring at a critical mass of about 6×1010M ⊙.A partial ex-planation may be a simple relation between mass and metal-licity:higher mass galaxies might have higher metallicity,and4Labbéetal.F IG .3.—The mass-to-light ratios in the rest-frame K −band (M /L K )of distant red galaxies (DRGs)and z =2−3Lyman break galaxies (LBGs).Symbols are as in Fig.1.(a )M /L K versus rest-frame U −V color,showing a clear correlation.On top we indicate the corresponding observed J s −K s color and the J s −K s >2.3limit (upward arrow ).Color tracks of an SSP (dotted line )and CSF model (dashed line )are shown.(b )The relation between M /L K and best-ﬁt age.The M /L K is more sensitive to the age of the stellar population than to dust extinction,hence the DRGs are on average older,not just more dusty versions of LBGs.(c )M /L K versus stellar mass.The highest-mass galaxies all have high M /L K ratios.The mass completeness limit (downward arrow )corresponds to our observed K s ,tot =22.5magnitude cut-off and a maximal M /L .A selection by rest-frame K-light corresponds to taking everything to the right of the constant L K line (dashed line ).Obviously this is still far from a vertical cut-off which would represent a selection by stellar mass.thus be redder.It is very noteworthy that there are “red and dead”galaxies at z =2−3.Previous authors found such galaxies at z =1−2(e.g.,Cimatti et al.2004;McCarthy et al.2004).Whereas the number density of “dead”galaxies at z =2.5is probably much lower than at z =0,the mere existence of these systems at such high redshift raises the question what caused such a rapid and early decline in star formation.Our model ﬁts imply that their star formation stopped by z =5or higher,close to or during the epoch of reionization.Possibly a strong feedback mechanism caused this,such as an AGN or galactic-scale outﬂow.We note that these galaxies are among the most massive galaxies at z =2.5,and hence were probably at the extreme tail of the mass function at z =5.Obviously,many questions remain unanswered by this study.Only a very small ﬁeld has been studied,and simi-lar studies on wider ﬁelds are necessary.Finally,the M /L ratios derived here remain model dependent and vary with the assumed SFH,metallicity,and IMF (e.g.,Bell et al.2003).Direct mass determinations are presently very challenging,but are becoming more important to understand the z =2−3galaxy population.This research was supported by the Carnegie Institution of Washington,the Netherlands Foundation for Research (NWO),the Leids Kerkhoven-Bosscha Fonds,the Lorentz Center,and the Smithsonian Institution.REFERENCESBarmby,P.,et al.2004,ApJS,154,97Bell,E.F.,et al.2003,ApJS,149,289Bruzual,G.&Charlot,S.2003,MNRAS,344,1000(BC03)Casertano,S.et al.AJ,120,pp.2747–2824,2000Calzetti,D.,et al.2000,ApJ,533,682Cimatti,A.,et al.2004,Nature,430,184Fazio,G.G.,et al.2004,ApJS,154,10Franx,M.et al.2003,ApJ,587,L79Förster Schreiber,N.M.,et al.2004,ApJ,616,40(F04)Kauffmann,G.,et al.2003,MNRAS,341,54Labbé,I.,et al.2003,AJ,125,1107Labbé,I.,2004,Doctoral Thesis,Leiden UniversityMadau,P.,et al.1996,MNRAS,283,1388McCarthy,P.J.,et al.2004,ApJ,614,L9Papovich,C.,Dickinson,M.,&Ferguson,H.C.2001,ApJ,559,620Rudnick,G.,et al.2003,ApJ,599,847Steidel,C.C.,et al.1996,AJ,112,352Steidel,C.C.,et al.1996b,ApJ,462,L17Stern,D.,et al.2005,ApJ,submitted (astro-ph/0410523)van Dokkum,P.G.et al.2003,ApJ,587,L83van Dokkum,P.G.,et al.2004,ApJ,611,703Yan,H.,et al.2004,ApJ,616,63。

Stellar populations in local star-forming galaxies. II.-Recent star formation properties an

a r X i v :a s t r o -p h /0209397v 2 20 D e c 2002Mon.Not.R.Astron.Soc.000,1–18(2002)Printed 2February 2008(MN L A T E X style ﬁle v2.2)Stellar populations in local star-forming galaxies.II.-Recent star formation properties and stellar masses.P.G.P´e rez-Gonz´a lez 1,A.Gil de Paz,5,2,1J.Zamorano,1J.Gallego,1A.Alonso-Herrero 3and A.Arag´o n-Salamanca 41Departamento de Astrof´ısica,Facultad de F´ısicas,Universidad Complutense,E-28040Madrid,Spain2NASA/IPACExtragalactic Database,California Institute of Technology,MS 100-22,Pasadena,CA 91125,USA3Steward Observatory,The University of Arizona,Tucson AZ 85721,USA4School of Physics and Astronomy,University of Nottingham,NG72RD,England5current address:The Observatories of the Carnegie Institution of Washington,813Santa Barbara St.,Pasadena,CA 91101,USAReceived 2February 2008ABSTRACTWe present the integrated properties of the stellar populations in the Universidad Complutense de Madrid (UCM)Survey galaxies.Applying the techniques described in the ﬁrst paper of this series,we derive ages,burst masses and metallicities of the newly-formed stars in our sample galaxies.The population of young stars is responsible for the H αemission used to detect the objects in the UCM Survey.We also infer total stellar masses and star formation rates in a consistent way taking into account the evolutionary history of each galaxy.We ﬁnd that an average UCM galaxy has a total stellar mass of ∼1010M ⊙,of which about 5%has been formed in an instantaneous burst occurred about 5Myr ago,and sub-solar metallicity.Less than 10%of the sample shows massive starbursts involving more than half of the total mass of the galaxy.Several correlations are found among the derived properties.The burst strength is correlated with the extinction and with the integrated optical colours for galaxies with low obscuration.The current star formation rate is correlated with the gas content.A stellar mass–metallicity relation is also found.Our analysis indicates that the UCM Survey galaxies span a broad range in properties between those of galaxies completely dominated by current/recent star formation and those of normal quiescent spirals.We also ﬁnd evidence indicating that star-formation in the local universe is dominated by galaxies considerably less massive than L ∗.Key words:galaxies:fundamental parameters –galaxies:evolution –galaxies:pho-tometry –galaxies:stellar content –infrared:galaxies –radio lines:galaxies1INTRODUCTIONThe present paper is the second of a series which deals with the determination of the main properties of the stellar pop-ulations in the Universidad Complutense de Madrid (UCM)Survey galaxies (Zamorano et al.1994,1996;Alonso et al.1999).We deal here with the integrated properties of the galaxies as a ﬁrst step towards understanding their evolu-tion.Future developments will address the properties of the spatially-resolved stellar populations and the improvement of the modelling procedures.This will be necessary to under-stand the details of the star formation history of each galaxy as well as their dust extinction properties,which turn out to be one of the key points (and probably the most important one)in this ﬁeld.One of the goals of this study is to determine the na-ture of the galaxies which were detected by the UCM Sur-vey.There is an extensive dataset available for the sample,including spectroscopic and photometric information cover-ing a broad wavelength range from the optical to the near infrared (nIR),together with some radio data.The analy-sis of the spectroscopic observations allow us to study the emission lines formed in the ionized gas clouds surrounding young hot stars.Among these lines,the Balmer H αline is one of the best tracers of the most recent star formation (Kennicutt 1992,1998a).It is easily observable in nearby galaxies and is less extinguished by dust than other optical emission lines (H β,[OII]λ3727˚A )and the ultraviolet contin-uum.The H αluminosity and equivalent width are directly linked to the youngest population of stars responsible for the heating and ionisation of the gas,and thus can be used in the determination of the mass in newly-formed stars,their2P.G.P´e rez-Gonz´a lez et al.age,etc.Spectroscopic data can also be used to evaluate the extinction(via the Balmer decrement),the metallicity,and the excitation.Photometric data covering a wide wavelength range can be used to carry out a population synthesis analysis of com-posite stellar populations.Many examples of such studies, for low and high redshift galaxies,are found in the litera-ture.See,e.g.,Kr¨u ger et al.(1995);de Jong(1996);Abra-ham et al.(1999);Brinchmann&Ellis(2000);Gil de Paz et al.(2000a);Bell&de Jong(2000);Papovich et al.(2001). Some other authors have focused on the quantitative anal-ysis of the optimal sets of observables and signal-to-noise ratios required to obtain robust results(see Bolzonella et al. 2000;Gil de Paz&Madore2002,and references therein).In this respect,the combination of high-quality optical, ultraviolet and nIR data has been found provides some of the fundamental information needed to study local galaxies. To complement the broad-band photometry,emission-line ﬂuxes can also be used in galaxies presenting star-formation activity.The ultraviolet part of the spectrum and the emis-sion lines are dominated by young hot stars formed recently. The nIR is essential to characterise the more evolved pop-ulation,since it is less sensitive to recent bursts and dust extinction.One principal application of this line of research is the determination of the stellar masses of galaxies,another ma-jor goal of our project.It has been argued that nIR data,and more precisely,the K-band luminosity,can be used as a good tracer of the stellar mass(Rix&Rieke1993;Brinchmann &Ellis2000).Based on this assumption,several nIR-based surveys have been carried out in order to use the K-band luminosity function at several redshifts to directly obtain the distribution galaxy masses(e.g.,Cowie et al.1996;Co-hen et al.1999;Kochanek et al.2001;Drory et al.2001). However,it is very important to test the reliability of the stellar masses determined using K-band luminosities alone. Age diﬀerences from galaxy to galaxy,or the presence of massive recent star-formation(with a mass comparable to that of the evolved population)may have an eﬀect on the mass-to-light ratio even in the nIR.Indeed,some authors have recently claimed that the K-band mass-to-light ratio depends on parameters such as the galaxy colours,clearly aﬀecting the determination of total stellar masses(Moriondo et al.1998;Brinchmann&Ellis2000;Bell&de Jong2001; Graham2002).P´e rez-Gonz´a lez et al.(2002b,Paper I hereafter)pre-sented the dataset and the modelling and statistical tech-niques used in the current analysis.Paper I also discusses how well our techniques are able to reproduce the ing a stellar population synthesis library,and taking into account the gas emission and dust attenuation, our method is able to model successfully the observational properties of star-forming galaxies.Several a priori parame-ters of the models were tested.These include(1)the evolu-tionary spectral synthesis library(we used Bruzual&Char-lot–private communication–and Leitherer et al.1999);(2) the recent star formation scenario(instantaneous and con-stant star formation rates-SFR-were tested);(3)the initial mass function(Salpeter1955,Scalo1986and Miller&Scalo 1979);and(4)the extinction-correction recipe(Calzetti et al.2000and Charlot&Fall2000).Among these,we found that the extinction plays a fundamental role.We present now the results obtained from the applica-tion of our modelling procedure and statistical analysis to the UCM Survey data.Brieﬂy,the global properties of the newly-formed stars and those of the underlying evolved pop-ulation will be quantiﬁed.These properties are derived for each individual galaxy,ensuring that the stellar content and star formation history of each object are properly taken into account.The determination of these properties will lead to a better understanding of the observational biases of this kind of surveys.A plan of the paper follows.First,the main proper-ties of the UCM Survey sample will be brieﬂy described in Section2.The population synthesis method used in this will be reminded in Section3(see Paper I for further de-tails).Next,the results concerning the youngest population will be presented and discussed in Section4.Following this, in Section5we will focus on the integrated stellar masses of the UCM galaxies.Finally,the conclusions will be pre-sented.Throughout this paper we use a cosmology with H0=70km s−1Mpc−1,ΩM=0.3andΛ=0.7.2THE SAMPLEThe present work has been carried out using the UCM Sur-vey sample composed of191galaxies selected by their Hαemission at an average redshift of0.026(Zamorano et al. 1994,1996;Gallego et al.1996).Within this sample,15ob-jects were classiﬁed as active galactic nuclei(AGN,including Sy1,Sy2and LINER types)by Gallego et al.(1996),and have been excluded from this study.Another11galaxies were observed in only two bands and the comparison with the models was not attempted.Theﬁnal sample is conse-quently formed by163galaxies(cf.Paper I).The extensive dataset used in this work includes optical and nIR imaging,and optical spectroscopy.For more details on the observations and the main spectroscopic and photo-metric properties of the galaxies,see Paper I and references therein.Although not presented in Paper I,in this paper we will also make use of the available HI21cm data for the UCM Survey galaxies.These data were obtained from the NASA/IPAC Extragalactic Database(NED).Most of the21 cmﬂuxes come from Huchtmeier&Richter(1989).We also used the data for11galaxies from Pisano et al.(2001).HI masses(in solar units)were calculated with the expression M HI=2.356·105·D2 Sdν,(1) where D is the distance in Mpc and Sdνis the integrated line-ﬂux in Jy km s−1(Roberts1975).3BASIC ASSUMPTIONS ANDMETHODOLOGYPaper I described the method to derive the properties of the most recent star-formation in star-forming galaxies using broad-band photometry and spectroscopy.Although the ba-sic assumptions and methodology were extensively described in that paper,we summarise them here to make this pa-per as self-contained as possible.The technique is based on the assumption that these galaxies have a composite stellarStellar populations in local star-forming galaxies.II.-Recent star formation properties and stellar masses.3Figure parison of the UCM objects (asterisks)with nor-mal relaxed (quiescent)spiral/irregular galaxies (triangles;Ken-nicutt 1998b).Vertical axis is plotted in solar units.population.The detection of nebular emission lines is un-doubtedly a hint for the presence of a very young stellar population,which will be referred as a recent burst of star formation ,the newly-formed stars or the recent starburst .This population is also responsible for the bluer colours ob-served in the UCM galaxies in comparison with ‘normal quiescent’(relaxed)spirals (see Alonso-Herrero et al.1996;P´e rez-Gonz´a lez et al.2000).This recent starburst is occurring in a spiral/lenticular galaxy.Morphological studies were carried out by Vitores et al.(1996)in the Gunn r band and P´e rez-Gonz´a lez et al.(2001)in the Johnson B band.A typical spiral intrinsically shows HII regions ionized by a population of recently-formed stars.Kennicutt (1983)and Davidge (1992)estimated the importance of this population (in comparison with the en-tire stellar content)measuring the H αequivalent width for a sample of normal spirals.On average,a typical relaxed Sb galaxy presents a value of 8˚A .However,the detection limit of the UCM Survey is EW (H α)∼20˚A (Gallego 1995).Consequently,the UCM objects must be experienc-ing a stronger burst of star formation in comparison with a normal galaxy.Figure 1shows the distinct nature of the UCM galaxies and normal spirals (Kennicutt 1998b).This plot depicts the stellar mass (as traced by the K -band luminosity)versus the luminosity from young stars (as traced by the H αlu-minosity).The latter has been normalized with the K -band luminosity in order to be able to compare between objects with diﬀerent luminosities/stellar masses.The nIR data for the comparison sample have been extracted from the Two Micron All Sky Survey (2MASS,Jarrett et al.2000),NED and de Jong &van der Kruit (1994).UCM galaxies appear as fainter objects than normal spirals but presenting larger normalized H αluminosities.This means that the present star formation is more important in comparison with the older population in the UCM objects than in normal spi-rals.Our modelling refers to the properties of a recent star formation event which takes place in excess of what is typi-cal in a normal spiral or lenticular galaxy .We have assumed that a recent burst of star formation (described by its age,metallicity and mass)is occurring in a galaxy whose colours and EW (H α)are those of a typical galaxy of the same morphological type.The assumed colours of the underlying evolved (older)population ,which have been taken from the literature,are the result of the past star formation history of these typical galaxies.The details of this past history are beyond the scope of this paper.Here we are not concerned with the detailed histories of individual galaxies,but with the statistical properties of our sample.The validity of this approach,at least in the statistical sense,is supported by the correlation between averaged colours and Hubble type for large samples of galaxies (see,e.g,Fukugita et al.1995;Fioc &Rocca-Volmerange 1999;Strateva et al.2001).In ad-dition,these underlying population colours are quite similar to our measurements in the outer parts of some randomly selected test galaxies (P´e rez-Gonz´a lez et al.2002a).The recent burst must be younger than ∼10Myr,since the EW (H α)drops considerably for ages older than this value.Given this short period of time,the star formation may be approximated by an instantaneous or constant SFR burst.A possible scenario with multiple bursts occurring all through the galaxy would be mimicked by a constant SFR model.This work also deals with the estimation of the total stellar mass of each galaxy.Mass-to-light ratios in the nIR (and in particular the K -band)have been claimed to be roughly independent of the galaxies’stellar populations and star-formation histories (Rix &Rieke 1993;Brinchmann &Ellis 2000).This statement will be discussed in Section 5.The modelling technique described in Paper I yields three parameters describing the population of newly-formed stars in the UCM galaxies.The three parameters are the age t ,metallicity Z and burst strength b (ratio between the mass of the starburst and the total stellar mass of the galaxy,i.e.,the importance of the recent star formation event).Both the observed colours and equivalent width that are ﬁtted by our modelling and the output parameters have been consid-ered as statistical distributions.The method also includes a Principal Component Analysis on the space of solutions which takes account of the degeneracies in this kind of stud-ies (cf.Paper I).The next sections will deal with the results obtained for these three properties.Paper I introduced and tested some input parameters that should be selected a priori .All of them refer to the stellar and nebular emission arising from the recent star-burst.As a reminder,we list here these parameters and the acronyms used hereafter:–The evolutionary synthesis model:Bruzual &Charlot (1999,private communication;BC99hereafter)or Leitherer et al.(1999,SB99from now on).–The star-forming mode of the young stellar popula-tion:instantaneous or continuous star formation rate.These modes will be referred to as INST and CONS.–The IMF:Salpeter (1955,SALP),Scalo (1986,SCA),or Miller &Scalo (1979,MSCA).In all cases,we use M low =0.1M ⊙and M up =100M ⊙for the lower and upper mass limits of the IMF.–The extinction recipe:Calzetti et al.(2000,CALZ00)or Charlot &Fall (2000,CF00).4P.G.P´e rez-Gonz´a lez etal.Figure 2.Distribution of the burst strengths derived for the UCM Survey galaxies.BC99instantaneous SFR models for the three initial mass functions and the CF00extinction recipe are plotted on the left and constant SFR models on the right. Each row corresponds to each one of the three IMFs considered: Salpeter(SALP),Scalo(SCA)and Miller-Scalo(MSCA).4PROPERTIES OF THE YOUNG STELLAR POPULATION4.1Burst strengthFig.2shows the histograms of the burst strength(in a log-arithmic scale)for the3IMFs considered and the CF00 extinction recipe.Left panels show results for the INST star formation and right panels for the CONS case.On the top of each diagram,the median value and quartiles are drawn. These quantities are also given in Table1,together with the relevant values for the rest of the derived quantities.The median values of the burst strengths are2–12%,depending on the input parameters of the models.The individual val-ues of b cover the whole range considered,from a pure young bursts(b≃1)to masses of new stars that are less than1% of the total mass of the galaxy.These results are similar to what was found for a smaller subsample in Gil de Paz et al. (2000a,hereafter GdP00),albeit with some of the galaxies studied now showing burst strengths close to100%.These high-b objects were not present in the subsample studied by GdP00.The models with SCA and MSCA IMFs present higher values of the burst strength than those using SALP by up to a factor of3–5.For the same age,the SCA modelsare parison of burst strengths for instantaneous and constant rates of star-formation.The points refer to the BC99 code,CF00extinction recipe and the3IMFs described in the text.redder than the MSCA ones,and these in turn are redder than the SALP ones.More young stars need to be added to the redder models in order to account for the observed colours,and the derived burst strength rises.The burst strengths derived using a constant SFR and an instantaneous burst are compared in Fig.3.The points scatter around the1–to–1line for all the IMFs.This indi-cates that,after a few million years,the observed proper-ties of a galaxy that experienced a massive instantaneous burst will resemble those of one experiencing a less eﬃ-cient but longer star formation event in which the mass of newly-formed stars is similar.We will come to this fact later. Galaxies show a small tendency to to have lower values of b for the CONS case than for the INST one.Indeed,for a given age,one can reproduce a certain EW(Hα)with a continuous burst less massive than an instantaneous one. The good agreement between the b values derived with the INST and CONS models for the galaxies with the lowest burst strengths is remarkable.These objects also show simi-larly young burst ages for both star-formation scenarios(see Section4.2).A comparison of the CONST and INST burst strengths derived for the SB99models with the CF00extinction law is presented in Fig.6,with very similar conclusions.On av-erage,SB99models show higher burst strengths than BC99 ones by a factor of0.1-0.2dex for case of CF00extinction and lower for the CALZ00prescription.In Paper I we concluded that the vast majority of the UCM galaxies are betterﬁtted with the INST models than with the CONS ones.Thus,the b values derived from the best-ﬁtting models will refer,in most cases,to the instanta-neous burst scenario.Only5objects have burst strengths higher than50% as derved from more than one model,including the model that best reproduces the observations.Four more joinStellar populations in local star-forming galaxies.II.-Recent star formation properties and stellar masses.5Figure 4.Distribution of the ages derived for the UCM Survey galaxies.BC99instantaneous SFR models for the three initial mass functions and CF00extinction are plotted on the left and constant star formation rate models on the right.this group if we only consider the best-ﬁtting model.Out of these 9objects,4are classiﬁed as SBN (Gallego et al.1996,see Paper I for a short description),two as DANS and three as HIIH objects.Most of them are com-pact objects (e.g.,UCM1256+2910,UCM2315+1923and UCM2319+2234),some have extended star-formation lo-cated throughout the object (as seen in H αimaging pre-sented in P´e rez-Gonz´a lez et al.2002c,e.g.,UCM0022+2049and UCM1306+3111).There are also two face-on galax-ies with clear spiral arms and a massive nuclear burst (UCM2256+2001and UCM2317+2356).All 9galaxies hav-ing stellar populations dominated by the young stars have relatively high extinctions,i.e.,E (B −V )=0.6−1.5.4.2AgeThe derived burst ages for the UCM galaxies show a rela-tively narrow peak at 5–6Myr for INST models (Fig.4,left).The median age and age distribution for the entire sample is almost independent of the IMF considered.The diﬀer-ence in mean age is only 0.01dex for the three IMFs.The independence of the derived burst ages on the IMF can be easily explained.The most important observable when de-termining the age of a young stellar population is EW (H α)(cf.Alonso-Herrero et al.1996,GdP00).The EW (H α)for young stellar populations is dominated by the most massivestars present,and thus the EW (H α)‘clock’only depends on the evolutionary clock of these massive stars,which doesn’t depend on the IMF.The reason for the narrow burst age distributions de-rived for the UCM galaxies can be explained from the way UCM galaxies were selected (see also discussion on Section 4.4).Only objects with relatively high H αequiv-alent widths (>20˚A )are present in the sample (Gallego 1995).Since EW (H α)drops sharply below that value after ∼10Myr (Fig.8),a sharp cutoﬀin the age distribution is expected at that age.The logarithmic nature of the x axis in Fig.4partially explains the drop in galaxy numbers for ages below 3Myr,since the time intervals encompassed by the low age bins is smaller.Moreover,for these very young ages the burst of star-formation is probably still hidden in very-high extinction regions (Gordon et al.1997),and there-fore galaxies with very young bursts will be hard to detect.The ages of the young stellar populations derived for the constant SFR models are not well constrained since EW (H α)changes very slowly with age (Alonso-Herrero et al.1996).The derived age distribution appears to be rather ﬂat for these models (Fig.4,right).The apparent excess of galaxies with older ages (log(t )∼8)is mainly due to the large time interval encompassed by the last bin.In any case,since the UCM galaxies clearly favour the INST models,the ages derived from the CONS models are largely irrelevant.When comparing BC99and SB99models (see Fig.6,middle-left panel),marginally younger burst ages (by 0.1dex)are found for the former,but the age distributions are similar.This behaviour derives directly from the fact that the predicted EW (H α)at any given age is higher in the SB99case.This is due to the diﬀerent evolutionary tracks and stellar libraries used in both sets of population synthesis models.4.3MetallicityAs discussed in GdP00and Paper I,the metallicity has a smaller eﬀect on the colours and EW (H α)’s predicted by the models than the burst strength and the age.Thus,the model-derived metallicities are much more uncertain.More-over,the population synthesis models used have metallici-ties with a small number of discrete values,which has a very strong eﬀect in the clustering of solutions in the parameter space (GdP00,Paper I).Although for the sake of complete-ness we will present in this section the metallicities derived by the models,extreme caution is needed when interpreting the results.Fig.5shows the distribution of model-derived metallic-ities for the young populations in the UCM sample galaxies.There is a large spread in the metallicities ﬁtted by our models,with more galaxies with metallicities below solar than above.The mean metallicity derived with the diﬀerent models is log(Z/Z ⊙) ≃−0.2.The derived metallicities are almost independent of the IMF considered.Finally,SB99models (Fig.6,lower panels)give lower metallicity values by 0.3–0.5dex,leaving less than 10%of the objects with metallicities above solar.In section 5we will discuss the spectroscopically-derived chemical abundance of the gas and its correlation with the galaxies’stellar masses.6P.G.P´e rez-Gonz´a lez et al.Table1.Median values and quartiles of the derived burst strengths,ages and metallicities.SALP CF00 6.71+0.13−0.117.45+0.50−0.59−1.53+0.33−0.39−1.80+0.60−0.53−0.16+0.33−0.27−0.04+0.32−0.36CALZ00 6.75+0.08−0.137.60+0.34−0.56−1.83+0.42−0.37−1.77+0.27−0.67−0.18+0.21−0.20−0.20+0.31−0.24SCA CF00 6.70+0.14−0.107.67+0.28−0.67−0.92+0.29−0.35−0.98+0.40−0.51−0.16+0.33−0.26−0.18+0.33−0.25CALZ00 6.74+0.08−0.147.46+0.36−0.46−1.24+0.31−0.30−1.21+0.32−0.54−0.16+0.19−0.20−0.24+0.23−0.22MSCA CF00 6.70+0.14−0.097.49+0.42−0.55−1.12+0.26−0.40−1.17+0.49−0.43−0.16+0.30−0.24−0.26+0.31−0.17CALZ00 6.75+0.08−0.137.22+0.44−0.32−1.54+0.34−0.31−1.51+0.41−0.54−0.17+0.16−0.21−0.20+0.23−0.23SALP CF00 6.79+0.11−0.117.92+0.07−0.45−1.36+0.41−0.42−1.60+0.45−0.33−0.54+0.44−0.44−0.55+0.38−0.53CALZ00 6.71+0.07−0.157.76+0.21−0.43−1.95+0.43−0.33−1.66+0.35−0.64−0.35+0.30−0.34−0.44+0.50−0.23Stellar populations in local star-forming galaxies.II.-Recent star formation properties and stellar masses.74.4CorrelationsFor the sake of simplicity,in this section and the reminder of this paper,all the plots will refer to the results obtained with the CF00extinction recipe,SB99models with instan-taneous SFR and a Salpeter IMF.As discussed in Paper I,this choice yields the best results when modelling the data,although the CALZ00extinction recipe seems to work marginally better for high extinction objects.In the plots, only galaxies with acceptableﬁts(as deﬁned in Paper I) will be shown.When relevant,results obtained with diﬀer-ent model parameter choices will also be mentioned in the discussion,including the set of results corresponding to the best-ﬁtting model for each galaxy.Fig.7shows the distribution of burst strengths accord-ing to the morphological type of each galaxy.Median values are indicated.The results for the SB99models suggest a relative modest increase in burst strength from Sa to Sc+. The number of galaxies classiﬁed as irregulars is too small to inferﬁrm results.Although this behaviour agrees with the idea that star formation is relatively more important in late spirals than in earlier ones,we remind the reader that our models assume an underlying stellar population in each galaxy similar to that of a‘normal’galaxy with the same morphological type.Thus,b refers to new stars formed in excess of what an average galaxy with the same morpholog-ical type would have(cf.Paper I).Another important point to remember when considering morphological trends is that the UCM sample is biased against low surface brightness ob-jects,since the galaxies were selected from objective-prism photographic plates.Moreover,for an S0galaxy to be de-tected in Hα,its star formation must be signiﬁcantly en-hanced with respect to a‘normal’S0.It is also worth point-ing that the relatively low burst strengths derived for Blue Compact Dwarf galaxies( b ≃5%)reveals the presence of an important underlying stellar population(Kr¨u ger et al. 1995;Gil de Paz et al.2000b,c;Kunth&¨Ostlin2000).Fig.8shows the relationship between model-derived age and EW(Hα)for the UCM galaxies.We use diﬀerent symbols for objects with diﬀerent log(b)values.Models for solar metallicity and several burst strengths(with an un-derlying population of a Sb galaxy,the most common mor-phology of the sample)have also been plotted.Both the models and the data show that for ages above∼2Myr, the EW(Hα)decreases with age.This is hardly surprising, since,as discussed above,EW(Hα)provides the strongest constraint when determining the ages.Given that at∼10 Myr the EW of the young stars equals the EW of the under-lying stellar population,the model predictions cross at that age.For older ages,the EW(Hα)for the composite stellar population becomes constant with time.For very low burst strengths(log(b)≃−3),EW(Hα)≃8˚A,i.e.,the value cor-responding to the underlying stellar population.For higher b values,EW(Hα)is dominated by the newly-formed stars, which have lower EW s.In Fig.9we draw the Hα/Hβratio(a measure of extinc-tion)versus the burst strength as derived with our method. Although a large scatter is present,the galaxies with the largest values of b seem to have typically larger extinctions, with Hα/Hβratios above5.0,i.e.,E(B−V)>0.5.Galax-ies with smaller bursts,on the other hand,seem to inhabit objects with lowerextinction.Figure7.Histograms of the burst strength for the UCM galaxies divided according to Hubble type(P´e rez-Gonz´a lez et al.2001). INT denotes interacting galaxies,and BCD Blue Compact Dwarf galaxies.Instantaneous SFR for SB99models,Salpeter IMF and CF00extinction is plotted on the left and the distribution of best-ﬁttings on theright.Figure8.Time evolution of the Hαequivalent width for solar metallicity models and diﬀerent burst strengths.Derived values for the UCM sample are also shown.Diﬀerent symbols are used for galaxies with diﬀerent log(b).Note that some of the points have log(b)values outside the range span by the models,but his is due to the fact that only solar metallicity models with aﬁxed underlying population are shown.If we had plotted the full range of models used,the models would encompass the full range of derived log(b)values(by construction).。

随着人越来越多我们要去别的星球住英语作文

随着人越来越多我们要去别的星球住英语作文As the population of Earth continues to explode, the question of whether we should seek new habitats on other planets becomes increasingly urgent. With finite resources and limited space, the Earth's ability to sustain humanlife is being stretched to its limits. In this essay, we will explore the reasons why interstellar migration may become necessary, the challenges we face in achieving it, and the potential benefits it could bring to humanity.Firstly, the population growth on Earth is exponential, with the global population expected to reach over 9 billion by 2050. This rapid growth is putting immense pressure on the planet's natural resources, leading to issues such as climate change, water scarcity, and food insecurity. As these problems worsen, the viability of sustaining human life on Earth becomes increasingly questionable.Secondly, the search for alternative habitats is motivated by the desire for human survival and progress. As a species, we have always been explorers and pioneers, pushing the boundaries of what is possible. Interstellarmigration would be a巨大leap forward in our evolutionary journey, allowing us to expand our horizons and potentially discover new sources of energy, minerals, and even life itself.However, the challenges of interstellar migration are numerous and daunting. The technology required to travel to other planets is still in its infancy, and the costs are astronomical. Additionally, the physical and psychological toll on astronauts during such long journeys is a major concern. Furthermore, the political and ethicalimplications of colonizing other planets are complex and controversial.Despite these challenges, the potential benefits of interstellar migration are immense. It could provide a new home for humanity, relieving the pressure on Earth's resources and environment. It could also lead to new scientific discoveries and technological advancements,推动人类文明的进步和发展。

stars1

• We will cover this in stellar evolution, but much of the heavy elements in our Universe today is created when a giant star end its life in a supernova（超新星）. The intense heat for this creates in the upper atmosphere （高层大气）of stars much of the elements we know - like iron, gold, even fluoride（氟化物） (yes, the same stuff in your toothpaste). • These elements disperse themselves to nearby molecular clouds. （分子云）When that cloud undergoes contraction（收缩） and give birth to a new star, the end result is a Population I star - one that is now metal rich. • One consequence of a metal rich star is that they are likely to contain a system of planets（行星）!
星团球状星团
、
open clusters.
疏散星团
Globular Clusters
They are generally comprised of old, metal poor stars that is contain little elements heavier than helium(氦). This means globular clusters formed a long time ago, before our galaxy contained more heavy stars.

1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。

a r
X
i
v
:a
s t
r o
-
p
h
/
3
1
1
2
2
2v
1
1
N
o
v
2
3
Dissertation Summary Stellar populations in local star-forming galaxies Pablo G.P´e rez-Gonz´a lez Current address:University of Arizona.Steward Observatory.933N Cherry Av.Tucson,AZ85721,USA Electronic mail:pgperez@ Thesis work conducted at:Departamento de Astrof´ısica,Facultad de Ciencias F´ısicas,Universidad Complutense de Madrid (Spain)Ph.D.Thesis directed by:J.Zamorano &J.Gallego ;Ph.D.Degree awarded:June 2003Received
The main goal of this thesis work1is studying the main properties of the stellar populations embedded in a statistically complete sample of local active star-forming galaxies:the Universidad Complutense de Madrid(UCM)Survey of emission-line galaxies 2.This sample contains191local star-forming galaxies at an average redshift of0.026.The survey was carried out using an objective-prism technique centered at the wavelength of the Hαnebular emission-line(a common tracer of recent star formation).
We have compiled an extensive multi-wavelength(from the optical to the near infrared) photometric(in narrow and broad bandﬁlters)and spectroscopic dataset for all the galaxies in the sample.The analysis of this dataset has been centered on the characterization of the morphology of the objects,on the comparison of the global properties of the sample with those of other galaxy samples at several redshifts,and on the study of the stellar content of the UCM galaxies.In the latter research topic,a special attention has been paid to the most recent star-forming bursts,which are the main characteristic of our objects.We have also developed(and tested with the dataset for the UCM sample)a powerful stellar population synthesis tool(which is able to obtain robust estimates of the stellar mass,a very important parameter in the study of the evolution of galaxies),and a technique to calculate luminosity and stellar mass functions.These techniques may be easily used in the characterization of samples of galaxies at any redshift.
The main results obtained in this work point out that the UCM Survey galaxies span a broad range in properties(among others,the total stellar mass and the star formation rate per stellar mass unit)between those of galaxies completely dominated by
pag
pag
Survey/index.html
current/recent star formation(e.g.,extreme dwarf HII galaxies or starburst galaxies)and those of‘normal’quiescent spirals.The UCM objects present intermediate-late Hubble type spiral morphologies and a recent instantaneous burst of star formation(occurred about5 Myr ago)with sub-solar metallicity.These bursts involved5%of the total stellar mass of the UCM galaxies.An‘average’UCM galaxy has a total stellar mass of1010solar masses, i.e.,about a factor of7-10lower than a typical local spiral galaxy.Finally,we estimate that13±3%of the total baryon mass density in the form of stars in the local Universe is associated with star-forming galaxies.。