The effect of helium-enhanced stellar populations on the ultraviolet-upturn phenomenon of early-type galaxies
aa r X i v : . [ a s t r o - ph . GA ] S e p THE EFFECT OF HELIUM-ENHANCED STELLAR POPULATIONS ONTHE ULTRAVIOLET-UPTURN PHENOMENON OF EARLY-TYPEGALAXIES
CHUL CHUNG, SUK-JIN YOON, AND YOUNG-WOOK LEE
Center for Galaxy Evolution Research and Department of Astronomy, Yonsei University, Seoul120-749, Korea; [email protected]
ABSTRACT
Recent observations and modeling of globular clusters with multiple populationsstrongly indicate the presence of super helium-rich subpopulations in old stellar sys-tems. Motivated by this, we have constructed new population synthesis models withand without helium-enhanced subpopulations to investigate their impact on the UV-upturn phenomenon of quiescent early-type galaxies. We find that our models withhelium-enhanced subpopulations can naturally reproduce the strong UV-upturns ob-served in giant elliptical galaxies assuming an age similar to that of old globular clustersin the Milky Way. The major source of far-UV (FUV) flux, in this model, is relativelymetal-poor and helium-enhanced hot horizontal branch stars and their progeny. TheBurstein et al. (1988) relation of the
F U V − V color with metallicity is also explainedeither by the variation of the fraction of helium-enhanced subpopulations or by thespread in mean age of stellar populations in early-type galaxies. Subject headings: galaxies: elliptical and lenticular, cD – galaxies: evolution — galaxies:stellar content – ultraviolet: galaxies — galaxies: individual (M87)
1. INTRODUCTION
It is well established that the far ultraviolet (FUV) flux (“UV-upturn”) observed in the nearbyquiescent early-type galaxies (ETGs) originates from a minority population of hot horizontal-branch(HB) stars and their progeny (see, e.g., Greggio & Renzini 1990; O’Connell 1999; Brown et al.2000). Recently, a significant progress has been made as to the origin of these hot HB starsin old stellar systems. This new theory is based on the recent observations and modeling ofglobular clusters (GCs) with extended HB, such as ω Cen and NGC 2808, where the multiple main-sequences and hot HBs in these clusters can only be explained by the presence of super helium-richsubpopulations (D’Antona & Caloi 2004; Norris 2004; Lee et al. 2005; Piotto et al. 2005, 2007).The origin of this helium enhancement is most likely due to the pollution from the intermediate-mass asymptotic giant branch stars and/or fast-rotating massive stars, or due to the enrichment bysupernovae (Ventura & D’Antona 2008; Decressin et al. 2007; Piotto et al. 2005; Lee et al. 2009). 2 –Therefore, the likely presence of super-helium-rich subpopulations and the resulting hot HB starsin old stellar systems deserve further investigation, as they could be a major source of the FUVflux in quiescent elliptical galaxies. The purpose of this Letter is to report our first result on theUV upturn phenomenon predicted from the Yonsei evolutionary population synthesis models withhelium-enhanced subpopulations.
2. POPULATION SYNTHESIS MODELS
The models presented in this paper were constructed using the Yonsei Evolutionary PopulationSynthesis (YEPS) code. The readers are referred to Park & Lee (1997), Lee et al. (2000), andChung et al. (2011, in prep. ) for the details of model construction. In order to investigate theeffect of helium-enhanced stellar population on the UV-upturn, we have used the most up-to-dateYonsei-Yale ( Y ) stellar isochrones and HB evolutionary tracks (Han et al. 2011, in prep. ) withdifferent values of helium abundance ( Y = 0 .
23, 0.33, and 0.38). We apply the same Reimers (1977)mass-loss parameter ( η ) and the enhancement in α -elements ([ α/F e ]) for both helium-enhancedand normal helium populations (see Table 1).Figure 1 shows an example of our model HR diagrams and corresponding spectral energydistributions (SEDs) of simple stellar population (SSP) for normal helium and helium-enhancedpopulations. Since helium-rich stars evolve faster than helium-poor stars, helium-rich stars havelower masses at given age. This effect has a striking difference on the HB stage (see left panels). Themean temperature of HB stars in helium-enhanced ( Y = 0 .
33) population is ∼ Y = 0 .
23) population at a given metallicity ([
F e/H ] = − . Y = 0 .
33 and0.38, as these values are roughly required to reproduce the extreme blue HB stars observed inthe Milky Way GCs (D’Antona & Caloi 2004, 2008; Lee et al. 2005; Piotto et al. 2005; Joo & Lee2011, in prep. ). In our modeling, populations having these two helium abundances were mixedhalf and half. Second, we have used mean
F U V − V colors of UV-strong GCs in M87 (Sohn et al.2006) to set the fraction of helium-enhanced subpopulations at given metallicity (see Figure 2a).Figure 2b shows the fraction of helium-enhanced subpopulations as a function of metallicity, whichwas adopted in our model with helium-enhanced subpopulations to reproduce the observed trendof F U V − V color with metallicity in M87 GCs. Figure 2c displays the contribution of helium-enhanced subpopulations in the metallicity distribution function (MDF) used in our constructionof the composite stellar population model for giant elliptical galaxies. This MDF was adopted fromthe simple chemical evolution model of Kodama & Arimoto (1997). 3 –According to the list of Lee et al. (2007), about 30% of the Milky Way GCs have extended HB,each of which contains about 30% of helium-enhanced subpopulation (Lee et al. 2005; Piotto et al.2005, 2007; Yoon et al. 2008; Han et al. 2009; Bellini et al. 2010), so about 9% of stellar populationin the Milky Way GCs are assumed to be helium-enhanced population. In our composite model forgiant elliptical galaxies, the number fraction of helium-enhanced subpopulations is about ∼ ∼ T eff ≥ ,
000 K) blue HB stars arecounted, which are actually responsible for the far-UV flux in our model giant elliptical galaxies.
3. COMPARISON WITH OBSERVATIONS
Figure 3 presents our composite models constructed with and without helium-enhanced sub-populations compared with the SEDs observed by
IUE satellite for two giant elliptical galaxiesNGC 4552 and NGC 4649. In our models, the value of mean metallicity ( h [ F e/H ] i ) was chosen sothat they can reproduce the observed Mg b ( ≈ α/F e ] is 0.3 in giant elliptical galaxies (Worthey et al. 1992; Thomas et al. 2005;Kormendy et al. 2009). The parameters adopted in our models presented in Figure 3 are listed inTable 1. It is clear from Figure 3 that our models with helium-enhanced subpopulations can natu-rally reproduce not only the strong far-UV upturns but also the 2500 ˚A dips observed in NGC 4552and NGC 4649, while the models without helium-enhanced subpopulations fail to reproduce the far-UV upturns, unless the age of underlying stellar population is increased to ∼
16 Gyrs (Park & Lee1997; Yi et al. 1999). Therefore, the minority population ( ∼ Our composite models with helium-enhanced subpopulations predict that about 85% of thefar-UV flux comes from the metal-poor side ([
F e/H ] ≤ .
0) of the MDF. In this regard, this modelis qualitatively similar to the “metal-poor HB model” of Park & Lee (1997), but it does not needto invoke unrealistically old ages ( ∼
16 Gyr) as in the model of Park & Lee (1997, see also Yi et al.1999). Our composite models with helium-enhanced subpopulations can reproduce the UV-upturnat the mean age of 11 Gyrs, in the age scale where the inner halo GCs of the Milky Way is 12 Gyrsold. If the age of the oldest GCs in the Milky Way is more like 13 Gyrs (Dotter et al. 2010), our agefor the giant elliptical galaxies would also increase by 1 Gyr. These ages for giant elliptical galaxies For the dwarf elliptical galaxy like M32, where the observed UV upturn is weak (
F UV − V = 7.3 in ABmagnitude system), all of the HB population is predicted to be cooler than 20,000 K. Brown et al. (2000) found fromtheir HST/STIS photometry of M32 that stars passing through the HB at T eff ≥ ,
500 K comprise only a smallfraction (about 7%) of the total HB population. This fraction is roughly consistent with that ( ∼ Most of the helium-rich HB stars in our models are too hot to have any significant effect on metal lines, such asCa II H + H ǫ /Ca II K index. The Ca II H + H ǫ /Ca II K index of our model for giant elliptical galaxy (presented inFigure 3) is 1.15, which is consistent with the observation ( ∼ ± b line strength against F U V − V color for the sample ofquiescent ETGs from Bureau et al. (2011). Superposed grids are our composite models with (blue)and without (red) helium-enhanced subpopulations. The observed ETGs show clear correlationbetween
F U V − V and Mg b that is analogous to the Burstein et al. (1988) relation. Our compositemodels predict that the strength of UV-upturn is controlled by the fraction of helium-enhancedsubpopulation, the mean age, and the mean metallicity of underlying stellar population. Amongthese three factors, the fraction of helium-enhanced subpopulation appears to be the most effectivefactor determining the strength of UV-upturn. For example, our 11 Gyr models with and withouthelium-enhanced subpopulations would explain most of the color spread in F U V − V . If all ETGsin this sample, however, contain similar fraction of helium-enhanced subpopulations, an age spreadspanning ∼ F U V − V color, in thesense that bluer galaxies are older. This is because the mean temperature of HB stars decreasesas the age of stellar population gets younger (see Lee et al. 1994; Park & Lee 1997). An agespread of this magnitude is also indicated from the Balmer line dating of ETGs (Trager et al. 2000;Thomas et al. 2005; Trager et al. 2008; Graves et al. 2009), which suggests that some age spreadis indeed responsible for the spread in F U V − V color. The observed fading of FUV flux withlook-back time for the bright cluster elliptical galaxies in z < .
4. DISCUSSION
We have demonstrated that the presence of relatively metal-poor and helium-enhanced subpop-ulations in ETGs can naturally reproduce the observed UV-upturn phenomenon, without invokingunrealistically old ages. Although the presence of helium-enriched stars appears to solve many ofthe problems associated with the UV-upturn phenomenon, it is important to note that the originof this helium enhancement is not fully understood yet. Consequently, it is not clear yet whetherthe proposed mechanisms to produce helium-enhanced stars in GCs will also be important in thegalactic scales. Nevertheless, several lines of evidence do suggest that a minority population ofhelium-enhanced stars would be present also in ETGs. First, the two well defined remaining localbuilding blocks in the Milky Way, the ω Cen and the Sagittarius dwarf galaxy (including M54),are all characterized by very extended HB with helium-enhanced extreme blue HB stars (Lee et al. Our models were constructed with [ α/F e ] = 0 .
3, while some galaxies in Figure 4 have lower values of [ α/F e ].If [ α/F e ] = 0 .
1, the
F UV − V color gets redder by small amount ( ∼ b index is more affected(decreases by ∼ α/F e ] decreases, and thisshould have only little effect on the predicted age spread among sample galaxies. in prep. ). Second, the orbital kinematics of Milky Way GCs with extended HB, which isindicating the presence of helium-enhanced subpopulation, are distinct from normal GCs, and arefully consistent with the hypothesis that they are the remaining relics of the early building blockspredicted in the hierarchical merging paradigm of galaxy formation (Lee et al. 2007; Bekki et al.2007). Finally, as discussed above, when the observed trend of F U V − V color with metallicityfor the UV-strong GCs (presumably with the extended HB) in the giant elliptical galaxy M87 iscombined with the expected MDF of a model giant elliptical galaxy, the F U V − V color similar tothat observed in giant elliptical galaxies is reproduced. This is again consistent with the buildingblock origin for these GCs, according to which the helium-enhanced subpopulations in ETGs wouldhave been supplied by these early galaxy building blocks.If normal ETGs are prevailed by helium-enhanced populations, as suggested in this Letter, theywould have impacts not only on UV, but also on other wavelength regime affected by relatively hotblue HB stars, such as Balmer absorption lines (Lee et al. 2000). In this respect, it is important tonote that Schiavon et al. (2006) found that their models (without helium-enhanced subpopulation)of passive evolution can not reproduce the unexpectedly strong Balmer line indices of ETGs at z ≈ . z ≈ .
9. It would be interesting to see how our models with helium-enhanced subpopulationsare matched with these observations. The well known discrepancy between the H β and higher orderBalmer lines (“Balmer mismatch”; Schiavon 2007) is also an important issue, for which the helium-enhanced populations could play a role. Our forthcoming paper (Chung et al. 2011, in prep. ) willdiscuss these issues in more detail, together with their connection with the UV-upturn phenomenon.We thank the referee for a number of helpful suggestions. Support for this work was providedby the National Research Foundation of Korea to the Center for Galaxy Evolution Research. SJYacknowledges support from Mid-career Researcher Program (No. 2009-0080851) and Basic ScienceResearch Program (No. 2009-0086824) through the NRF of Korea, and support from the KASIResearch Fund 2011. REFERENCES
Bekki, K., Yahagi, H., Nagashima, M., & Forbes, D. A. 2007, MNRAS, 382, L87Bellini, A., Bedin, L. R., Piotto, G., Milone, A. P., Marino, A. F., & Villanova, S. 2010, AJ, 140,631Brown, T. M., Bowers, C. W., Kimble, R. A., Sweigart, A. V., & Ferguson, H. C. 2000, ApJ, 532,308Bureau, M., et al. 2011, MNRAS, 414, 1887Burstein, D., Bertola, F., Buson, L. M., Faber, S. M., & Lauer, T. R. 1988, ApJ, 328, 440 6 –D’Antona, F., & Caloi, V. 2004, ApJ, 611, 871D’Antona, F., & Caloi, V. 2008, MNRAS, 390, 693Decressin, T., Meynet, G., Charbonnel, C., Prantzos, N., & Ekstr¨om, S. 2007, A&A, 464, 1029Dorman, B., O’Connell, R. W., & Rood, R. T. 1995, ApJ, 442, 105Dotter, A., et al. 2010, ApJ, 708, 698Graves, G. J., Faber, S. M., & Schiavon, R. P. 2009, ApJ, 693, 486Greggio, L., & Renzini, A. 1990, ApJ, 364, 35Han, S.-I., Lee, Y.-W., Joo, S.-J., Sohn, S. T., Yoon, S.-J., Kim, H.-S., & Lee, J.-W. 2009, ApJ,707, L190Kodama, T., & Arimoto, N. 1997, A&A, 320, 41Kormendy, J., Fisher, D. B., Cornell, M. E., & Bender, R. 2009, ApJS, 182, 216Lee, H.-c., Yoon, S.-J., & Lee, Y.-W. 2000, AJ, 120, 998Lee, J.-W., Kang, Y.-W., Lee, J., & Lee, Y.-W. 2009, Nature, 462, 480Lee Y.-W., Demarque P., Zinn, R., 1994, ApJ, 423, 248Lee, Y.-W., et al. 2005, ApJ, 621, L57Lee, Y.-W., Gim, H. B., & Casetti-Dinescu, D. I. 2007, ApJ, 661, L49Norris, J. E. 2004, ApJ, 612, L25O’Connell, R. W. 1999, ARA&A, 37, 603Park, J.-H., & Lee, Y.-W. 1997, ApJ, 476, 28Piotto, G., et al. 2005, ApJ, 621, 777Piotto, G., et al. 2007, ApJ, 661, L53Ree, C. H., et al. 2007, ApJS, 173, 607Reimers, D. 1977, A&A, 57, 395Rose, J. A. 1985, AJ, 90, 1927Schiavon, R. P., et al. 2006, ApJ, 651, L93Schiavon, R. P. 2007, ApJS, 171, 146 7 –Siegel, M. H., et al. 2007, ApJ, 667, L57Sohn, S. T., O’Connell, R. W., Kundu, A., Landsman, W. B., Burstein, D., Bohlin, R. C., Frogel,J. A., & Rose, J. A. 2006, AJ, 131, 866Thomas, D., Maraston, C., Bender, R., & Mendes de Oliveira, C. 2005, ApJ, 621, 673Trager, S. C., Faber, S. M., Worthey G., Gonzalez J. J. 2000, AJ, 119, 1645Trager, S. C., Faber, S. M., & Dressler, A. 2008, MNRAS, 386, 715Ventura, P., & D’Antona, F. 2008, MNRAS, 385, 2034Worthey, G., Faber, S. M., & Gonzalez, J. J. 1992, ApJ, 398, 69Yi, S., Demarque, P., & Oemler, A., Jr. 1998, ApJ, 492, 480Yi, S., Lee, Y.-W., Woo, J.-H., Park, J.-H., Demarque, P., & Oemler, A., Jr. 1999, ApJ, 513, 128Yoon, S.-J., Joo, S.-J., Ree, C. H., Han, S.-I., Kim, D.-G., & Lee, Y.-W. 2008, ApJ, 677, 1080
This preprint was prepared with the AAS L A TEX macros v5.2. −2−1 0 1 2 3.6 3.8 4 4.2 4.4 L og L / L Log T eff[Fe/H] = −0.9 Y = 0.33 −2−1 0 1 2 L og L / L [Fe/H] = −0.9 Y = 0.23 −3−2−1 0 1 0 2000 4000 6000 L og f l − L og f V l (Å) MS to RGBHBPAGB Total Flux −3−2−1 0 1 L og f l − L og f V MS to RGBHBPAGB Total Flux
Fig. 1.— The synthetic HR diagrams and corresponding spectral energy distributions for simplestellar populations. Upper panels are models for normal helium abundance ( Y = 0 . Y = 0 .
33) population. The abbreviations MS, RGB, HB,and PAGB refer the flux contributions from main-sequence, red giant branch, horizontal branch(including post-HB progeny), and post asymptotic giant branch stars, respectively. 9 – -2.5-2-1.5-1-0.5 0 0.5 3 4 5 6 7 8 [ F e / H ] (FUV-V)
47 TucNGC 6388NGC 6441 w Cen M54 (a)
Milky Way GCsM87 GCsw/o He-rich Populationw/ He-rich Population F r ac ti on o f S t a r s (b) Normal He Pop.He-rich Pop. R e l a ti v e F r e qu e n c y [Fe/H] Normal He Pop.He-rich Pop.MDF for model ETG(<[Fe/H]> = -0.10, [ a /Fe] = 0.3) (c) Fig. 2.— ( a ) Comparison of our models ( solid lines ) with GCs in M87 ( circles ) and the Milky Way( triangles ). The data are from Sohn et al. (2006) for GCs in M87 and from Dorman et al. (1995) for GCsin the Milky Way, and the F U V − V color is in the AB magnitude system. Bared arrows are upper limitsfor the metal-rich GCs in M87. The red line is a model with only normal helium population ( Y = 0 . t = 12 Gyr), while the blue line is a model when the helium-enhanced subpopulations ( Y = 0 .
33 + 0.38; t = 11 Gyr) are added to the normal helium population ( Y = 0 . t = 11 Gyr) using the fraction inFigure 2b. Some well-known Milky Way GCs with extended HB ( ω Cen, M54, NGC 6388, and NGC 6441),together with 47 Tuc, are labeled. [
F e/H ] values for ω Cen and M54 are weighted mean values of multiplepopulations in these GCs. ( b ) The fraction of helium-enhanced population calibrated to reproduce F U V − V colors of UV-strong GCs in M87 at given metallicity. In our models (the blue line in Figure 2a), helium-enhanced subpopulations are added to the normal helium populations with this relation. ( c ) The metallicitydistribution function (Kodama & Arimoto 1997) adopted in our composite model for giant elliptical galaxyand the contribution from helium-enhanced subpopulations. Considering the metallicity spread of compositemodel peaked at [ F e/H ] = 0 .
3, the actual contribution from helium-enhanced subpopulation is only ∼
10 – −2−1 0 0 1000 2000 3000 4000 5000 6000 L og f l − L og f V l (Å) <[Fe/H]> = −0.10 [ a /Fe] = 0.30 w/o He−rich pop.w/ He−rich pop.NGC 4649 −2−1 0 L og f l − L og f V <[Fe/H]> = −0.10 [ a /Fe] = 0.30 w/o He−rich pop.w/ He−rich pop.NGC 4552 Fig. 3.— The comparison of observed SEDs of NGC 4552 and NGC 4649 (data from Burstein et al.1988 and Yi et al. 1998) with our composite models. The composite models presented here are basedon the series of SSP models such as those displayed in Figure 1, with the metallicity distributionfunction in Figure 2c. Our models with helium-enhanced subpopulation can reproduce the observedfar-UV upturns at the age of 11 Gyrs. 11 – M g b (FUV - V) < Z > Age
11 Gyr 8 Gyr 7 Gyr 6 Gyr 5 Gyr14 Gyr 12 Gyr 8 GyrN4552 w/o He-rich populationw/ He-rich population
Fig. 4.— The (
F U V − V ) color vs. Mg b correlation for the sample of quiescent early-typegalaxies having H β ≤ h Z i and age (Gyr) are indicated. 12 –Table 1. INPUT PARAMETERS ADOPTED IN OUR COMPOSITE STELLARPOPULATION MODEL FOR NGC 4552 Parameters Helium-enhanced population Normal Helium populationInitial mass function Salpeter ( x = 1 .