Radial Dependence of the Proto-Globular Cluster Contribution to the Milky Way Formation
Chul Chung, Mario Pasquato, Sang-Yoon Lee, Ugo N. di Carlo, Deokkeun An, Suk-Jin Yoon, Young-Wook Lee
aa r X i v : . [ a s t r o - ph . GA ] S e p Draft version September 5, 2019
Typeset using L A TEX twocolumn style in AASTeX62
Radial Dependence of the Proto-Globular Cluster Contribution to the Milky Way Formation
Chul Chung, Mario Pasquato,
2, 3
Sang-Yoon Lee, Ugo N. di Carlo,
Deokkeun An, Suk-Jin Yoon, andYoung-Wook Lee Department of Astronomy & Center for Galaxy Evolution Research, Yonsei University, Seoul 03722, Republic of Korea INAF, Osservatorio Astronomico di Padova, vicolo dell’Osservatorio 5, 35122 Padova, Italy INFN, Sezione di Padova, Via Marzolo 8, I35131, Padova, Italy Dipartimento di Scienza e Alta Tecnologia, University of Insubria, Via Valleggio 11, I22100, Como, Italy Department of Science Education, Ewha Womans University, 52 Ewhayeodae-gil, Seodaemun-gu, Seoul 03760, Republic of Korea (Accepted for publication in the ApJL September 1, 2019)
ABSTRACTRecent interpretation of the color − magnitude diagrams of the Milky Way (MW) bulge has suggestedthat the observed double red-clump feature can be a natural consequence of He-enhanced stellar pop-ulations in the MW bulge. This implies that globular clusters (GCs), where the He-enhanced second-generation (SG) stars can be efficiently created, are the most likely candidate contributors of He-richstars to the MW bulge. We extend this idea to the Galactic inner halo and investigate the fraction ofthe SG stars as a function of the Galactocentric distance. We use bluer blue-horizontal branch (bBHB)stars, which are assumed to be originated from He-rich SG populations, as proxies of SG stars, and findthat the fraction of bBHB stars increases with decreasing Galactocentric distance. Simulations of theGC evolution in the MW tidal field qualitatively support the observed trend of bBHB enhancement inthe inner halo. In these simulations, the increasing tidal force with decreasing Galactocentric distanceleads to stripping of stars not only from the outskirts but also from the central regions of GCs, whereSG stars are more abundant. We discuss the implication and prospect of our findings concerning theformation history of the bulge and inner halo of the MW. Keywords: globular clusters: general — stars: abundances — stars: evolution — stars: horizontal-branch — methods: numerical — Galaxy: halo INTRODUCTIONUnderstanding the formation history of the MilkyWay is one of the ultimate goals of galaxy formationstudies (e.g., Somerville & Dav´e 2015; Conselice 2014).Various simulations of galaxy formation have revealedthat, in addition to the stars that originated in-situ ,the most plausible candidates for the building blocks ofthe Milky Way are massive dwarf galaxies and/or mas-sive globular clusters (GCs; e.g., Lada & Lada 1991).Numerous studies on identifying the primary buildingblocks of the Milky Way have been carried out, butthe nature of the building block that played a domi-nant role in building up the stellar populations of theMilky Way is still under debate. Especially the Galac-
Corresponding author: Chul Chung, Mario [email protected], [email protected] tic halo has been considered as a collection of smallerstructures such as dwarf spheroidals, and it has beenwell established that dwarfs are the major constituentsof the Milky Way halo via accretion (e.g., Searle & Zinn1978; Klimentowski et al. 2007). Compared to the dwarfgalaxies, the contribution of GCs as building blocksof the Milky Way has been regarded as small becauseof their lower current mass without dark matter, aswell as the small number of GCs (Tolstoy et al. 2009;Deason et al. 2015).In this context, the discovery of He-enhanced popu-lations in the Milky Way bulge, however, changed thecurrent paradigms on the formation history of the MilkyWay. Proto-GCs, where He- and N-enhanced starscan be formed (Martell & Grebel 2010; Martell et al.2016; Schiavon et al. 2017), are the most likely can-didates for the birthplace of bulge He-enhanced starsthrough the self-enrichment of the first generation pol-
Chung et al. luters (Kim & Lee 2018), and the observed double redclumps (RCs) in the bulge (resulting from He-rich stars)are crucial evidence for their contribution to the bulge(Lee et al. 2015; Joo et al. 2017; Lee et al. 2019). Therecent study on chemical tagging of the double RC inthe bulge further strengthens this view by detectingN-enhanced stars in the bright RC (Lee et al. 2018),which implies that almost 50% of stars in the bulgemay be originated from second-generation (SG) stars.This is in line with a picture of a rapid orbital decayof proto-GCs emerging from N-body models of MilkyWay evolution, which suggests that the nuclear star clus-ter observed in the Galactic center is the result of re-peated GC mergers (e.g., Capuzzo-Dolcetta & Miocchi2008; Abbate et al. 2018). Given the possibility of thesignificant fraction of GC-origin population in the bulgeregion, one would expect that the effect of proto-GCson the formation of the halo should be non-negligible.Chung et al. (2016) demonstrated the preferential re-moval of first-generation (FG) stars from GCs in theouter halo and cautiously suggested that this processmay result in the misunderstanding that most of theouter halo consists of the dwarf-origin FG stellar pop-ulations. On the other hand, the observed RC starsin the bulge (Lee et al. 2018) may be the simple out-come of full or almost full disruption of GCs at smallGalactocentric distances. Therefore, for intermediateGalactocentric distances, one would expect a partialstripping of SG stars, depending on the interplay be-tween the strength of tidal stripping and the amountof confinement of the SG stars in the center of theGC, which in turn depends on the details of SG forma-tion (D’Ercole et al. 2008; H´enault-Brunet et al. 2015;Bastian & Lardo 2018) and dynamical evolution withinthe host GC (Vesperini et al. 2013; Dalessandro et al.2014; Lim et al. 2016). This may lead to a varying frac-tion between FG and SG stars in the bulge and halofields.In this study, we investigate how proto-GC systems,which usually have experienced multiple generations ofstar formation, have influenced the formation of theMilky Way inner halo. We analyze the number ratiobetween FG and SG stars as a function of the Galacto-centric distance using the ratio of bluer blue horizontalbranch (bBHB) to redder blue horizontal branch (rBHB)stars as proxies for each of the populations, respectively.We find a gradient of the bBHB / rBHB ratio with Galac-tocentric distance, which we interpret in terms of in-creasing preferential tidal stripping of FG stars as Galac- Throughout this paper, the SG stars indicate all generationsof stellar populations enriched by FG stars in proto-GCs. tocentric distance increases. This is expected in mas-sive GCs, where relaxation did not have enough time tomix the different stellar generations and SG stars arestrongly confined to the central regions of the GC (e.g.,see Vesperini et al. 2013). To interprete our results, weadditionally run simplified dynamical evolution simula-tions based on the numerical solution scheme for Hillequations proposed by Quinn et al. (2010) for GCs atdifferent Galactocentric distances.The paper is organized as follows. Section 2 describesthe construction of a population synthesis model andthe sample selection. In Sections 3 and 4, we presentour new findings together with results from simulations.Section 5 discusses the implications of our results for theformation history of the Milky Way. POPULATION SYNTHESIS MODEL ANDSAMPLE SELECTIONThe evolutionary population synthesis (EPS) modelspresented here are constructed under the same assump-tions on the input parameters adopted in Chung et al.(2016). Readers are referred to Chung et al. (2013) andChung et al. (2017) for detailed prescriptions of the EPSmodels.Figure 1 presents synthetic horizontal branch (HB)stars and corresponding isochrones from the main se-quence to the tip of red giant branch stars in ( u − g ) and ( g − r ) color–magnitude diagrams (CMDs) at givenmetallicities and 12 Gyr age. Based on these CMDs, wecarefully selected blue colored areas where only bBHBsare located in both colors of 0 . < ( u − g ) < . − . < ( g − r ) < − .
25. For rBHB stars, we adoptedthe selection criteria of Deason et al. (2015). To avoidRR Lyrae contamination in the sample, we took starswith ( g − r ) < − . g ) of bBHBs within blueboxes in ( u − g ) and ( g − r ) colors are 0.49 and 0.70 m ag, respectively. Given that the mean M g of rBHBsin the blue box is 0.43 m ag, the M g differences amongselection boxes is less than 0.3 m ag, which can be usedas distance indicator at a given magnitude. From topto bottom panels, the metallicity of both FG and SGincreases and bottom panels show the case of inner halometallicity and age. This metallicity range clearly showsthe drastic morphology change of HB stars originatedfrom SG populations at the age of 12 Gyr.We make use of SDSS photometry (York et al. 2000)from DR14 for the halo star census. Our candidate starsare selected from Photoprimary . We are mainly con-cerned with halo rBHB and bBHB stars at Galactocen-tric distances from 4 to 50 kpc, which corresponds to theapparent g magnitude of rBHB stars ranging between 15 adial Dependence of Proto-GC Contribution to MW Formation mag . We have trimmed the stars below b = 30 ◦ to avoid bulge or disk stars , and selected our samplewith photometric errors less than 0.05. To see the effectof the population shift in the same halo region with re-spect to the Galactocentric distance, we choose a rathernarrow volume-limited star sample in − ◦ < l < ◦ to meet the above criteria.We use the color–color diagram (CCD), which is notaffected by the distance, to select a clean sample of starsin the same evolutionary stages, i.e., rBHBs and bBHBs,in the halo. Figure 2 shows halo stars selected from theSDSS DR14 in the ( u − g ) vs. ( g − r ) plane. Weoverplot models of synthetic HB stars as well as youngturn-off stars on top of the selected halo stars to showhow rBHBs and bBHBs are distributed in the CCD.The position of these model HB stars on the CCD areindicated as solid lines with blue and cyan colors for rB-HBs and bBHBs, respectively. The adopted metallicitiesfor HB stars in the diagram are the same as models inthe bottom panels of Figure 1. In order to show whereblue stragglers (BSs) are placed in the CCD, we providethe dashed lines of young main-sequence turn-off pointswhich cross the lower part of the area. Blue and reddashed lines are for the metallicities of [Fe / H] = − . − .
0, respectively. rBHB stars and some RR Lyraesfrom metal-poor populations are placed in the curvedhook-shaped area in the CCD and sweep the upper edgeof the hook-shape. We intend to select rBHBs and bB-HBs simultaneously, so we narrowed down our sampleusing the guidelines of our population model. Amongthe selected sample, we use 5,942 stars that satisfy ourcriteria for the analysis in Section 3. The final selectionboxes with all samples in the density map are presentedin the right panel of Figure 2. The high-density regionsin the CCD are shown as dark black colors. THE FRACTION OF THESECOND-GENERATION POPULATIONS INTHE INNER HALOFigure 3 shows the selection boxes of rBHBs and bB-HBs with respect to the magnitude bin. The bottom leftpanel shows the location of our volume-limited samplesin the heliocentric coordinate. To analyze the popu-lation shift with respect to the Galactocentric distance,we divided halo stars into four regions, and those regionsare indicated by numbers in the plot. As the magnitudeand the Galactocentric distance increase, stars in the The GAIA-Enceladus candidates exist everywhere in the halobut are well-known to have much more overlaps with the thickdisk (e.g., Helmi et al. 2018). By selecting stars with b > ◦ and not including solar neighborhood stars, we have reduced thecontamination caused by the GAIA-Enceladus. bBHB selection box gradually decrease. Following thebBHB/rBHB selection boxes, the fraction of bBHB-to-rBHB decreases with the increasing Galactocentric dis-tance from 1 to 4 region, implying that the fraction ofbBHBs increases with the decreasing Galactocentric dis-tance. The bottom right panel of Figure 3 demonstratesthis trend with respect to the Galactocentric distance.There exists an apparent contrast in bBHB/rBHB ra-tios between the region inside 10 kpc and the regionoutside. This is consistent with earlier studies based onB-V colors (e.g., Preston et al. 1991) and spectroscopicselection of BHB stars (Santucci et al. 2015).This result is particularly interesting because it iswell established that metallicity of stellar componentsof Galactic halo increases as Galactocentric distancedecreases (e.g., Layden 1994; Torrealba et al. 2015),which in turn result in more red HBs in the innerhalo. Also, the age estimations of Galactic inner halo inthe literature suggest around ∼ . / H] ∼ − . , we need to adopt He-rich populations usually observed in GCs. Since He-richstars evolve faster than normal He stars, the He-richstellar populations originated from GCs could reproducemore bBHB stars at younger ages with a high metallic-ity of inner halo (Chung et al. 2011, 2017). Figure 1displays this effect by showing the population synthesis The age and CNO abundance also affect the HB morphol-ogy. However, considering the increasing N abundance (e.g.,Martell et al. 2011; Schiavon et al. 2017) and small age differencewith decreasing Galactocentric distance, bBHBs are most likelyoriginated from He-rich populations.
Chung et al. model of simple stellar populations under the differentassumptions on the initial He abundance. BHBs are nat-urally reproduced even at the intermediate metallicity of[Fe / H] ≃ − . DYNAMICAL EVOLUTION MODELS OFPROTO-GLOBULAR CLUSTERSThe observed trend can be explained by a coupleof galaxy formation scenarios. One of such possibili-ties includes a gradual age shift (Santucci et al. 2015;Carollo et al. 2016). In this Letter, we propose that itcan also be plausibly explained by disruption of mas-sive GCs by tidal force, and subsequent ejection of He-enriched SG stars with systematically bluer colors.We use our C implementation of the symplectic,time-reversible second order integrator introduced byQuinn et al. (2010) to compute the orbits of stars ina Plummer potential (representing the GC) in a circu-lar orbit around a point-mass galaxy (representing theMilky Way). The forces acting on each star include tidesand the Coriolis force, which, being velocity-dependent cannot be treated correctly with simpler integrationschemes such as standard leapfrog (Rein & Tremaine2011).We generate stars with an initially multivariate nor-mal distribution with standard deviation equal to a/ √ x , y , z coordinates, where a is the Plummerscale radius, and equal to v c / √ v x , v y , v z where v c is the equilibrium circular velocity at the scaleradius. We then tag as SG the stars located within dis-tance ka from the GC center, where k is allowed to varyover different simulations from 0 . ≈
10% and 60%. Theorbits are evolved from these initial conditions for 1000crossing times and considered escapers if their final posi-tion is further than 100 a from the center of the GC. Overour set of simulations, we change the ratio of the revo-lution time (around the Galactic center) to the crossingtime by several orders of magnitude. This correspondsto changing the ratio of the Plummer scale radius to theGalactocentric distance.We find two expected trends in the results of our sim-ulations: first, the fraction of escapers at the end of thesimulation is higher in GCs that are nearer to the Galac-tic center with respect to those that are further awayfrom it. This is clearly a direct effect of tidal stripping.Second, we find that SG stars that are confined withina sphere around the center in the initial conditions ofour simulations are less likely to be escapers than theaverage star. This latter result holds at any Galacto-centric distance but is more pronounced far from theGalactic center, where the Hill sphere of the cluster islarger, and stars located centrally are even more unlikelyto escape. In Figure 4, we plot the fraction of SG es-capers over FG escapers as a function of Galactocentricdistance. Simulations with different initial ratios of SGstars over the total number of stars are shown in differ-ent colors. The behavior is essentially the same, withalmost full stripping at low Galactocentric radii corre-sponding to a fraction of SG stars among the escapersequal to the initial fraction. As Galactocentric distanceincreases, only the stars that were initially unbound andthe outermost parts of the GC are stripped, so that com-paratively fewer SG stars escape. DISCUSSION AND CONCLUSIONWe have found a population gradient in the inner haloof the Milky Way as shown by the decreasing fractionof bBHB-to-rBHB stars with increasing Galactocentricdistance based on SDSS photometry. He-rich SG stel-lar populations originated from proto-GCs can explainthis trend if the fraction of SG tidal stripping increasedas Galactocentric distance decreases. We provide sim- adial Dependence of Proto-GC Contribution to MW Formation . The closest example M31halo also has enhanced α -elements which are similarto the abundance patterns of the Milky Way halo(Vargas et al. 2014). Proto-GCs are enriched in α -elements fed by SNe type II and they were disrupted toform the halo of the Milky Way rather than present-daydwarf satellites. This may be yet another manifestationthat the primary building blocks of the inner halo areproto-GCs.Interestingly, recent abundance analyses of early-typegalaxies support our predictions of the increasing SGpopulation with the decreasing Galactocentric distance.As shown e.g., in Milone et al. (2017) and generally re-garded as well established, the enhancement of severalelements, such as N and Na, would usually accompanyHe enhancement. van Dokkum et al. (2017) (Figure 10)directly confirmed our hypothesis of the galaxy forma-tion by showing enhanced Na abundance and depleted Oabundance with decreasing radius of early-type galaxies,which are the same behavior of elements originated fromGCs. If the dominant fraction of SGs causes this result,the assertion related to the bottom-heavy initial massfunction, usually suggested and supported by strong Naabsorptions (e.g., van Dokkum & Conroy 2010), is notvalid, but the effect of SG stars originated from GCs ismore viable. We will discuss this issue in the upcomingpaper regarding the initial mass function of early-typegalaxies (Chung et al. in prep.).ACKNOWLEDGMENTSWe thank the referee for helpful suggestions. C.C.,S.-J.Y., D.A., and Y.-W.L. acknowledge support pro-vided by the National Research Foundation (NRF) ofKorea to the Center for Galaxy Evolution Research (No.2017R1A2B3002919, and 2017R1A5A1070354). S.-J.Y. Other competing ideas suggest that massive dwarf galax-ies built the inner halo (e.g., Battaglia et al. 2017; Deason et al.2018), which must have experienced stronger dynamical friction.Their masses are also large enough that they could contain alarge number of α -enhanced stars originated from asymptotic gi-ant branch stars. Chung et al. acknowledges support from the Mid-career ResearcherProgram (No. 2019R1A2C3006242). M.P. acknowl-edges financial support from the European Unions Hori-zon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No. 664931.U.N.D.C. acknowledges financial support from Univer-sit`a degli studi dell’Insubria through a Cycle 33rd PhDgrant.REFERENCES
Abbate, F., Mastrobuono Battisti, A., Colpi, M., et al.2018, Monthly Notices of the Royal AstronomicalSociety, 473, 927Bastian, N., & Lardo, C. 2018, Annual Review ofAstronomy and Astrophysics, 56, 83Battaglia, G., North, P., Jablonka, P., et al. 2017,Astronomy & Astrophysics, 608, A145Capuzzo-Dolcetta, R., & Miocchi, P. 2008, Monthly Noticesof the Royal Astronomical Society: Letters, 388, L69Carollo, D., Beers, T. C., Placco, V. M., et al. 2016, NaturePhysics, 12, 1170Cava, A., Schaerer, D., Richard, J., et al. 2018, NatureAstronomy, 2, 76Chung, C., Lee, Y.-W., & Pasquato, M. 2016, MonthlyNotices of the Royal Astronomical Society: Letters, 456,L1Chung, C., Yoon, S.-J., Lee, S.-Y., & Lee, Y.-W. 2013, TheAstrophysical Journal Supplement, 204, 3Chung, C., Yoon, S.-J., & Lee, Y.-W. 2011, 740, L45—. 2017, The Astrophysical Journal, 842, 91Conselice, C. J. 2014, Annual Review of Astronomy andAstrophysics, 52, 291Dalessandro, E., Massari, D., Bellazzini, M., et al. 2014,The Astrophysical Journal Letters, 791, L4Deason, A. J., Belokurov, V., Koposov, S. E., & Lancaster,L. 2018, The Astrophysical Journal Letters, 862, L1Deason, A. J., Belokurov, V., & Weisz, D. R. 2015,Monthly Notices of the Royal Astronomical Society:Letters, 448, L77D’Ercole, A., Vesperini, E., D’Antona, F., McMillan, S.L. W., & Recchi, S. 2008, Monthly Notices of the RoyalAstronomical Society, 391, 825Dessauges-Zavadsky, M., Schaerer, D., Cava, A., Mayer, L.,& Tamburello, V. 2017, The Astrophysical JournalLetters, 836, L22Elmegreen, B. G., Bournaud, F., & Elmegreen, D. M. 2008,The Astrophysical Journal, 688, 67Hawkins, K., Jofr´e, P., Gilmore, G., & Masseron, T. 2014,Monthly Notices of the Royal Astronomical Society, 445,2575Helmi, A., Babusiaux, C., Koppelman, H. H., et al. 2018,Nature, 563, 85 H´enault-Brunet, V., Gieles, M., Agertz, O., & Read, J. I.2015, Monthly Notices of the Royal AstronomicalSociety, 450, 1164Jofr´e, P., & Weiss, A. 2011, Astronomy & Astrophysics,533, A59Joo, S.-J., Lee, Y.-W., & Chung, C. 2017, TheAstrophysical Journal, 840, 98Kalirai, J. S. 2012, Nature, 486, 90Kim, J. J., & Lee, Y.-W. 2018, The Astrophysical Journal,869, 35Klimentowski, J., Lokas, E. L., Kazantzidis, S., et al. 2007,Monthly Notices of the Royal Astronomical Society, 378,353Lada, C. J., & Lada, E. A. 1991, In: The formation andevolution of star clusters (A93-48676 20-90), 13, 3Layden, A. C. 1994, The Astronomical Journal, 108, 1016Lee, Y.-W., Hong, S., Lim, D., et al. 2018, 862, L8Lee, Y.-W., Joo, S.-J., & Chung, C. 2015, Monthly Noticesof the Royal Astronomical Society, 453, 3906Lee, Y.-W., Kim, J. J., Johnson, C. I., et al. 2019, TheAstrophysical Journal Letters, 878, L2Lim, D., Lee, Y.-W., Pasquato, M., Han, S.-I., & Roh,D.-G. 2016, The Astrophysical Journal, 832, 99Martell, S. L., & Grebel, E. K. 2010, Astronomy &Astrophysics, 519, A14Martell, S. L., Shetrone, M. D., Lucatello, S., et al. 2016,The Astrophysical Journal, 825, 146Martell, S. L., Smolinski, J. P., Beers, T. C., & Grebel,E. K. 2011, Astronomy & Astrophysics, 534, A136Milone, A. P., Piotto, G., Renzini, A., et al. 2017, MonthlyNotices of the Royal Astronomical Society, 464, 3636Preston, G. W., Shectman, S. A., & Beers, T. C. 1991,Astrophysical Journal, 375, 121Quinn, T., Perrine, R. P., Richardson, D. C., & Barnes, R.2010, The Astronomical Journal, 139, 803Rein, H., & Tremaine, S. 2011, Monthly Notices of theRoyal Astronomical Society, 415, 3168Santucci, R. M., Beers, T. C., Placco, V. M., et al. 2015,The Astrophysical Journal Letters, 813, L16Schiavon, R. P., Zamora, O., Carrera, R., et al. 2017,Monthly Notices of the Royal Astronomical Society, 465,501Searle, L., & Zinn, R. 1978, Astrophysical Journal, 225, 357 adial Dependence of Proto-GC Contribution to MW Formation Somerville, R. S., & Dav´e, R. 2015, Annual Review ofAstronomy and Astrophysics, 53, 51Tolstoy, E., Hill, V., & Tosi, M. 2009, Annual Review ofAstronomy and Astrophysics, 47, 371Tolstoy, E., Venn, K. A., Shetrone, M., et al. 2003, TheAstronomical Journal, 125, 707Torrealba, G., Catelan, M., Drake, A. J., et al. 2015,Monthly Notices of the Royal Astronomical Society, 446,2251 van Dokkum, P., Conroy, C., Villaume, A., Brodie, J., &Romanowsky, A. J. 2017, The Astrophysical Journal,841, 68van Dokkum, P. G., & Conroy, C. 2010, Nature, 468, 940Vargas, L. C., Gilbert, K. M., Geha, M., et al. 2014, TheAstrophysical Journal Letters, 797, L2Vesperini, E., McMillan, S. L. W., D’Antona, F., &D’Ercole, A. 2013, Monthly Notices of the RoyalAstronomical Society, 429, 1913York, D. G., Adelman, J., Anderson, J. E. J., et al. 2000,The Astronomical Journal, 120, 1579
Chung et al. -1 0 1 2 3 4 5 6 [Fe/H]=-1.1[Fe/H]=-1.0 M g FGSG-1 0 1 2 3 4 5 6 [Fe/H]=-0.8[Fe/H]=-0.7 M g FGSG-1 0 1 2 3 4 5 6 0 0.4 0.8 1.2 1.6 2 2.4[Fe/H]=-1.4[Fe/H]=-1.3 M g (u-g) -0.4 0 0.4 0.8 1.2FGSG(g-r) Figure 1.
The synthetic color − magnitude diagrams (CMDs) of proto-GCs with FG and SG populations. The left and rightpanels are ( u − g ) and ( g − r ) CMDs at 12 Gyr, respectively. Red and blue colors depict the FG and SG populationsat given metallicities, respectively. The metallicities of SGs are set as 0.1 dex higher than those of FGs. Crosses between − . ≤ ( g − r ) ≤ . . ≤ ( u − g ) ≤ . adial Dependence of Proto-GC Contribution to MW Formation −0.4 −0.3 −0.2 −0.1 0(g−r) ( u − g ) ( u − g ) (g−r) b > 30°| l | < 15° 4 kpc < d GC < 50 kpcRR LyraerBHB BS and Young TOsbBHB Figure 2.
Left: volume-limited halo stars selected from SDSS DR14
PhotoPrimary . The position of different evolutionarystages of stars in the ( u − g ) and ( g − r ) diagram are indicated based on HB stars presented in the bottom panels of Figure 1.Blue solid, cyan solid, and blue line with crosses indicate rBHB, bBHB, and RR Lyrae stars, respectively. Blue and red dashedlines for BSs and young turn-off stars are predicted by simple stellar population models of [Fe / H] = − . − .
0. Right: thedensity map of halo stars and selection boxes of rBHB and bBHB stars in the color–color diagram. Chung et al. g < 1630 (cid:176) < b < 60 (cid:176) ( u - g ) (g-r) -0.4 -0.3 -0.2 -0.1 0bBHB/rBHB = 0.1415 < m g < 1660 (cid:176) < b < 90 (cid:176) (g-r) -0.4 -0.3 -0.2 -0.1 0bBHB/rBHB = 0.0816 < m g < 1760 (cid:176) < b < 90 (cid:176) (g-r) -0.4 -0.3 -0.2 -0.1 0bBHB/rBHB = 0.1017 < m g < 1960 (cid:176) < b < 90 (cid:176) (g-r) b = 30 (cid:176) (cid:176) -15 (cid:176) < l < 15 (cid:176) GC( kpc ) x z ( kpc )-8-16-24-32-40-48 81624324048 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 10 100 b B H B / r B H B Galactocentric distance (kpc)
Figure 3. ( Upper panels ) The CCD for measurements of the number ratio between rBHBs and bBHBs in the Milky Way halowith respect to the Galactocentric distance. The point size of stars is inversely proportional to the photometric error. The rBHBmagnitude (M g ∼ .
45) from 15 to 19 corresponds to the Galactocentric distance from 4.0 to 50.0 kpc. The number of starswithin the bBHB selection box is drastically reduced as the Galactocentric distance increases. (
Bottom left ) The volume-limitedhalo stars in the heliocentric plane. The red dot marks the center of the Milky Way. The denoted numbers on the plot indicatethe location of the samples presented above. (
Bottom right ) The trend of the bBHB-to-rBHB ratio with the Galactocentricdistance. Dark and pale grey boxes indicate the cases of | l | < ◦ and | l | < ◦ , respectively. For presentation purposes, a slightextension in the Galactocentric distance was introduced for the broader l case in the 1 region. The width of lines indicates therelative errors caused by uncertainty in the number of stars within the selection box. Even with a larger sample, the resultshows an almost similar trend to the narrow l case. adial Dependence of Proto-GC Contribution to MW Formation
100 200 500 1000 . . . . . Galactocentric distance / Plummer scale radius S G e sc ape r s / F G e sc ape r s Initial SG fraction30 %40 %50 %