1. Stellar chemistries in dwarf Spheroidal galaxies and the Magellanic Clouds Vanessa Hill 1.Stellar chemistries - chemical evolution: 1.Commonalities and peculiarities in dSph (and LMC) 2.Rôle of AGBs 3.Extremely low metallicity stars in dSph

2. The standard picture (seminal) Matteucci & Brocato 1990

3. Star formation histories Tolstoy, Hill & Tosi ARAA 2009 NOW B-R Sculptor (m-M) 0 = 19.7 Carina (m-M) 0 = 20.0 Monelli et al. 09 Hildago et al. 09 Cole et al. 07 Matteo priv com. De Boer et al. 2010 and 2011

4. DART samples M V =-11.1 M V =-13.2 M V =-9.3M V =-9.5 Koch et al 2006, 2008

5. SculptorFornax (V-I) Zaritsky LMC Bar Disk Sgr: (Bellazzini et al. 2006) 12% old (>10Gyrs) >80% 5.5-9Gyrs ~6% younger stars (BP) Sagittarius Tolstoy et al. 2004, Battaglia et al. 2008 Battaglia et al. 2006 DART FLAMES Letarte et al. 2010 Lemasle et al. in prep Tolstoy Hill Tosi 2009 DART Venn et al. 2011, Lemasle et al. 2011 Battaglia et al. 2010 see talk by A. McWilliam Van Der Swaelmen et al. 2013 & PhD: 160 *

6. The standard picture  -elements SNII: [  /Fe] ~0.4 (fast enrichment) SNIa:: Fe, no  (delayed enrichement) Milky-Way Venn et al. 2004 SNII+SNIa knee

7. Tolstoy, Hill, Tosi 2009 ARAA Sgr Sbordone et al. 2007 + McWilliam et al. 2005 Milky-Way Venn et al. 2004 Distinct evolution

8. Tolstoy, Hill, Tosi 2009 ARAA Sgr Sbordone et al. 2007 + McWilliam et al. 2005 Fornax Letarte et al. 2010 Milky-Way Venn et al. 2004

9. Distinct evolution Each galaxy occupies a different locus - evolutionary track [  /Fe] « knee » metallicity: according to the ability of the galaxy to retain metals –Sgr : [Fe/H]knee >-1.2 –Fnx: [Fe/H]knee <-1.5? (Lemasle et al. in preparation for an outer field cover nicely - 1.0<[Fe/H]<-2.7 dex) –Scl: [Fe/H]knee ~ -1.8 in accordance to total L of the galaxy reflected on mean metallicity linked to SFH (gas availablility?) Dispersion: none detected in Scl, Fnx, Sgr Abundance pattern in the metal-poor stars everywhere undistinguishable ? -> IMF Tolstoy, Hill, Tosi 2009 ARAA Sgr Sbordone et al. 2007 + McWilliam et al. 2005 Fornax Letarte et al. 2010 Sculptor Tolstoy Hill Tosi 2009 & Hill et al. in prep) + Shetrone et al. 2003& Geisler et al. 2005 Milky-Way Venn et al. 2004 SNII +SNIa

10. Distinct evolution In the common metallicity regime, clusters and field stars agree. alpha elements are depleted compared to the MW disk, as expected from: –slower SF (e.g. Matteucci & Brocato 1990, Pagel & Tautvaisiene1998) –Bursts of star formation (Wyse 1996, Pagel & Tautvaisiene1998) –Galactic winds (e.g. Freitas Pacheco 1998; Lehner 2007 evidence of outflow) The position of the « knee » in [  /Fe] is ill-defined, because of the lack of metal- poor field stars… Old and metal-poor GCs resemble the galactic halo. No evidence for a different massive star IMF LMC field Pompeia, Hill et al. 2008, Van der Swaelmen, Hill et al. 2012 clusters Hill et al. 2000, and 2009; Johnson et al. 2006; Mucciarelli et al. 2008, 2010 SMC cluster Hill et al. in prep; * * RR Lyrae stars Haschke et al. 2012 Milky-Way Venn et al. 2004

11. Is there truely a knee in Scl ? Kirby et al. 2010 MRS data: sorted by errors

12. Is there truely a knee ? Kirby et al. 2010 MRS data: sorted by errors Texte DART data

13. There is a true plateau (and a knee) Kirby et al. 2010 MRS data: sorted by errorsDART data Low Z stars in Scl: Starkenburg et al. 2012 (Xshooter); Tafelmayer et al. 2010 (UVES) + 1 star Frebel et al. 2010

14. Distinct evolution Dispersion: probably present in Sextans (at all metallicities ?) -> inhomogeneous Sgr Sbordone et al. 2007 + McWilliam et al. 2005 Fornax Letarte et al. 2010 Sculptor Tolstoy Hill Tosi 2009 & Hill et al. in prep) + Shetrone et al. 2003& Geisler et al. 2005 Sextans Kirby et al. 2010 +Shetrone et al. 2003+Aoki et al 2009 Milky-Way Venn et al. 2004 Each galaxy occupies a different locus - evolutionary track [  /Fe] « knee » metallicity: according to the ability of the galaxy to retain metals Sgr : [Fe/H]knee >-1.2 Fnx: [Fe/H]knee <-1.5? (Lemasle et al. in preparation for an outer field cover nicely -1.0<[Fe/H]<-2.7 dex) Scl: [Fe/H]knee ~ -1.8 Sextans: knee ????? in accordance to total L of the galaxy (& mean metallicity)

15. Distinct evolution Dispersion: Detected in Carina -> bursty SFH + inhomogeneous Probably present in Sextans (at all metallicities ?) -> inhomogeneous At the lowest metallicities (EMPS), inhomogeneities or smooth as in MW halo ? Carina Venn et al. 2011, Lemasle etal. 2011 +Koch et al. 2008 + Shetrone et al. 2003 Milky-Way Venn et al. 2004 Each galaxy occupies a different locus - evolutionary track [  /Fe] « knee » metallicity: according to the ability of the galaxy to retain metals Sgr : [Fe/H]knee >-1.2 Fnx: [Fe/H]knee <-1.5? Scl: [Fe/H]knee ~ -1.8 Carina: no knee, dispersion ! in accordance to total L of the galaxy (& mean metallicity)

16. Sph code Nbody-Tree-SPH code with simple chemistry (Mg, Fe): cosmologically motivated initial conditions, isolated galaxies, feedback treated with care. varying Mtot, ρ g, r max, c *, (ε SN,t ad ) reproduces L-metallicity and M/L-L relations Revaz, Jablonka (2009, 2012)

17. Modelling Scl in a cosmological contexte Romano & Starkenburg 2013

18. The standard picture  -elements SNII: [  /Fe] ~0.4 (fast enrichment) SNIa:: Fe, no  (delayed enrichement) Neutron-capture element R- process: massive stars (fast enrichment) S-process: AGB stars (slower enrichement) Milky-Way Venn et al. 2004 SNII+SNIa r-process +s-process knee r-process +SNIa

19. N-capture elements 2nd peak r- and s- process elements (Ba, La,..) similar to MW in Scl: –Dominated by r- process at the lowest metallicities (all galaxies) –Mix of r- and s- process [Fe/H]> -2 In Fornax, Sgr and the LMC, the s- process displays a strong enhancement at the highest metallicities (younger ages): AGBs s-process yields are metallicity- dependent (seeds), favoring high-A over low-A elements (Ba/La over Y/Zr), and Y or Zr are observed not to be enhanced in Fnx/Sgr/LMC.  Points towards low-metallicity AGB pollution  Models: Lanfranchi et al. (dSph); Tsujimoto & Bekki (LMC)  Should be seen in Carbon ? r- and s- Sgr Sbordone et al. 2007, + McWilliam et al. 2005 Fornax Letarte et al. 2010 Sculptor Tolstoy Hill Tosi 2009 & Hill et al. in prep) + Shetrone et al. 2003& Geisler et al. 2005 Milky-Way Venn et al. 2008 SS r- LMC LMC field Van der Swaelmen, Hill et al. 2012 clusters Hill et al. 2000, and 2009; Johnson et al. 2006; Mucciarelli et al. 2008, 2010 * * RR Lyrae stars Haschke et al. 2012

20. Europium The “pure” r-process element Eu (>93% in the Sun): should behave like an  elements Eu follows  as expected in Scl On the contrary, Eu (and but [Eu/  ]) is significantly enhanced in the metal-rich part of Sgr, Fnx and the LMC, i.e. in the galaxies with also strong AGB contributions  Tempting to suggest an AGB contribution to Eu…. (also CEPM- sr stars). But see A. McWilliam’s talk. Sgr McWilliam et al. 2005 Fornax Letarte et al. 2010 Sculptor Tolstoy Hill Tosi 2009 & Hill et al. in prep) + Shetrone et al. 2003& Geisler et al. 2005 Milky-Way Venn et al. 2008 LMC field Van der Swaelmen, Hill et al. 2012 clusters Hill et al. 2000, and 2009; Johnson et al. 2006; Mucciarelli et al. 2008, 2010 * * RR Lyrae stars Haschke et al. 2012

21. Sculptor: bringing all together Fitting wide field & deep CMD simultaneously with the MDF: beats the age-metallicity degeneracies overall fit of CMD (x2) not necessarily better, but builds a self-consistent picture T. de Boer et al. 2011(DART)

22. Sculptor: bringing all together T. de Boer et al. 2011(DART)

23. Sculptor: spatial variations T. de Boer et al. 2011(DART)

24. Sculptor: spatial variations Sculptor is not only case (Sextans: Battaglia et al. 2010,...) linked to environment ? (seems not to be seen in isolated dSph)  ! Caution when interpreting MDFs derived in the center only (bias towards younger, metal-rich) Tolstoy et al. 2004 Battaglia et al. 2008a, 2008b

25. Sculptor: bringing all together The SFH can conversely be used to constrain solutions to derive more robust ages on the RGB The SNIa (knee) in Scl occurred 2Gyrs after the start of SF Presumably no SNIa pollution will be observed in Scl outskirts (no data yet). A self-consistent model (reproducing SFH, chemical enrichment timescale) for the two populations (incl. kinematics) is yet to be produced. (e.g. Revaz & Jablonja 2011 could not reproduce population gradients) T. de Boer et al. 2011

26. Quantifying the number of low-Z stars C alibrated to the low-Z regime, CaII triplet survey (DART) yields metallicities down to -4. The shape of the low-Z tail is undistiguishable from that of the MW halo ([Fe/H]<- 2.5) (Kirby 2008) Starkenburg et al. 2010

27. Low-Z tail: Characterising low-Z stars Low-Z follow-up ( ): ‣ Subaru (Aoki et al. 2009) ‣ VLT/UVES (Tafelmeyer et al. 2010: Fnx, Scl, Sext) ‣ VLT/Xshooter (Starkenburg et al. in prep, Scl) ‣ Magellan/MIKE (Venn et al. 2011 subm., Car) Stars with [Fe/H]<<-3 confirmed in dSph Their chemical composition are mostly similar to those of the MW halo Hints of true dispersion Shetrone et al. 2001, 2003 Koch et al. 2008, 2009 Cohen &Huang 2009 Frebel et al. 2009, 2010 Tolstoy, Hill, Tosi 2009

28. Summary Chemical evolution consistent with a less efficient enrichment (SN Ia contributions at lower Fe than in MW) ÙThe position of the knee seems to correlate with the galaxy L or or Mv ÙExpected if lowest mass systems loose easily their gas (less efficient enrichment) ÙWhen star formation is extremely low & bursty, dispersion may prevails at all metallicities (Carina?, Sextans?) Very high s-process element content of the metal-rich populations in Fornax, Sagittarius and the LMC (but NOT in Sculptor, Carina, Draco), result from a strong pollution by metal-poor AGBs. dSph do host some extremely metal poor stars (<-3), although in small numbers Metal poor stars abundances are not (yet) very significantly different from those of the metal-poor halo (but see M. Shetrone’s talk about C) Many modelling efforts ongoing….

30. Summary 1.Star formation histories: the last 10 years have relied on a CDMs from shallow ground-based large-coverage surveys (e.g. Zaritsky et al. ) plus a few very deep CMD in tiny fields (HST, e.g. Dolphin et al. 2001 ). We have now entered an era of HST plus ground-based, very deep and large fov imaging (eg. Gallart et al. 2008, Noel et al. 2007, 2009 ). See talks by M. Cignoni, R. Carrera. 2.Metallicity distributions/age-metallicity relations: for a long time, clusters were the dominant probes. Large spectroscopic surveys are providing reliable metallicity distributions (from individual stars) in the field. Gradients are still an open issue. See talks by W. Hankey, R. Carrerra… 3.Detailed abundances and chemical evolution: large samples of of stars in the LMC disk and bar (field and clusters) show: a.LMC disk chemical evolution consistent with slower evolution (SN Ia contributes at lower Fe than in the MW disk) b.Very abundant s-process elements traced to a strong pollution by metal-poor AGBs. c.Numerous nucleosynthetic puzzles opened up…See talk of M. Van der Swaelmen d.The metal-poor (and EMP) population remains elusive in current surveys e.SMC remains essentially uncharted territory

31. Manganese models: ‣ different SFH: ‣ Scl/Scl, Fnx/Fnx ‣ without metallicity- dependent SNIa yields: ‣ ‣ with metallicity-dependent SNIa yields: North et al. (DART) 2012

32. 1- From CMDs to SFHs Tolstoy, Hill, Tosi 2009 ARAA

33. 3- Detailed elemental abundances Zaritsky LMC ~50kpc OGLE SMC ~70kpc Present gas (HII regions) Young population ( 12 (SGs, …) Intermediate ages (1-10 Gyr) V>16-17 (AGBs, RGBs) Old population ( >10Gyr) V>16 (RGBs), V>19 (RR Lyr)

34. LMC NIR (2MASS) H I 7.5deg Zaritsky et al. 2004 Young populations: Irregular Older populations: smooth SFH, especially at young ages, will depend on field !

35. Metallicity distribution: LMC Bar Outer disk Carrera et al. 2008a Cole et al. 2005 LMC Bar LMC: No strong gradient of RGBs along the disk The bar may be slightly more metal-rich (- 0.3 vs -0.5). “halo” (similar to old GCs) present in all regions around [Fe/H]=-1.5. Similar to outer halo (Majevski 2009) ? Cole et al. 2009

36. Metal-poor stars ? Random RGB stars samples in both clouds fail to sample properly the metal-poor tails (if any). RR Lyrae select old (and presumably metal-poor stars) LMCSMC Haschke et al. 2012 (OGLE III RRLyr)

37. Carrera 2008b Age-metallicity relations LMC SMC Cole et al. 2005 Carrera 2008a

38. Age-metallicity relations Sharma et al. 2010Parisi et al. 2009 SMC LMC

39. N-capture elements r- and s- Pure r- r- and s- LMC field Pompeia, Hill et al. 2008, Van der Swaelmen, Hill et al. 2012 clusters Hill et al. 2000, and 2009; Johnson et al. 2006; Mucciarelli et al. 2008, 2010 SMC cluster Hill et al. in prep; * * RR Lyrae stars Haschke et al. 2012 Milky-Way Venn et al. 2004 s- process very efficient in LMC disk, with a take off around [Fe/H]>-1.0 –s-process yields are metallicity- dependent (seeds), favoring high-A over low-A elements (Ba over Y).  Points towards low-metallicity AGB pollution r- to s- transition can be used as clock (delayed AGB products): around [Fe/H]=-1.0dex  LMC AMR implies ~10 Gyrs at this metallicity  First cluster to show s-process is 9Gyrs old

40. N-capture elements r process: should behave like an  elements. On the contrary, Eu is enhanced in the LMC (and SMC) compared to the MW disk. (already known in supergiants, Russell&Bessell 1989, Luck & Lambert 1992, Hill et al. 1995, 1999, Luck et al. 1998 ) s- process very efficient in LMC disk, with a take off around [Fe/H]>-1.0 –s-process yields are metallicity- dependent (seeds), favoring high-A over low-A elements (Ba over Y).  Points towards low-metallicity AGB pollution r- to s- transition can be used as clock (delayed AGB products): around [Fe/H]=-1.0dex  LMC AMR implies ~10 Gyrs at this metallicity  First cluster to show s-process is 9Gyrs old r- process LMC field Pompeia, Hill et al. 2008, Van der Swaelmen, Hill et al. 2012 clusters Hill et al. 2000 and 2009; Johnson et al. 2006; Mucciarelli et al. 2008, 2010 SMC cluster Hill et al. in prep; * * RR Lyrae stars Haschke et al. 2012 Milky-Way Venn et al. 2004 r- process

41. Distinct evolution In the common metallicity regime, clusters and field stars agree. alpha elements are depleted compared to the MW disk, as expected from: –slower SF (e.g. Matteucci & Brocato 1990, Pagel & Tautvaisiene1998) –Bursts of star formation (Wyse 1996, Pagel & Tautvaisiene1998) –Galactic winds (e.g. Freitas Pacheco 1998; Lehner 2007 evidence of outflow) The position of the « knee » in [  /Fe] is not very defined, even in the LMC: lack of metal-poor field stars… This hampers a proper timing of the early chemical enrichment. Old and metal-poor GCs resemble the galactic halo. Field stars statistics are exceedingly low… No evidence for a different massive star IMF LMC field Pompeia, Hill et al. 2008, Van der Swaelmen, Hill et al. 2012 clusters Hill et al. 2000, and 2009; Johnson et al. 2006; Mucciarelli et al. 2008, 2010 SMC cluster Hill et al. in prep; * * RR Lyrae stars Haschke et al. 2012 Milky-Way Venn et al. 2004

42. N-capture elements r- and s- Pure r- r- and s- LMC field Pompeia, Hill et al. 2008, Van der Swaelmen, Hill et al. 2012 clusters Hill et al. 2000, and 2009; Johnson et al. 2006; Mucciarelli et al. 2008, 2010 SMC cluster Hill et al. in prep; * * RR Lyrae stars Haschke et al. 2012 Milky-Way Venn et al. 2004 r process: should behave like an  elements. On the contrary, Eu is enhanced in the LMC (and SMC) compared to the MW disk. (already known in supergiants, Russell&Bessell 1989, Luck & Lambert 1992, Hill et al. 1995, 1999, Luck et al. 1998 ) s- process very efficient in LMC disk, with a take off around [Fe/H]>-1.0 –s-process yields are metallicity- dependent (seeds), favoring high-A over low-A elements (Ba over Y).  Points towards low-metallicity AGB pollution r- to s- transition can be used as clock (delayed AGB products): around [Fe/H]=-1.0dex  LMC AMR implies ~10 Gyrs at this metallicity  First cluster to show s-process is 9Gyrs old

43. LMC Pompeia, Hill et al. 2008 Sgr Sbordone et al. 2007 Fornax Letarte et al. 2010 Milky-Way Venn et al. 2008 Similarities within the local group… LMC, Sgr, Fornax: extremely similar chemical peculiarities ( , odd elements, iron peak, s- process) Dominated by intermediate-age population Totally different gaz content, ongoing star formation …different orbits… Sgr most tidally disturbed, Fnx less, LMC even less (more massive)  What about the SMC ?

44. LMC Pompeia, Hill et al. 2008 Sgr Sbordone et al. 2007 Fornax Letarte PhD 2007 Sculptor Tolstoy Hill Tosi 2009 & Hill et al. in prep Milky-Way Venn et al. 2008 … and distinct evolution Scl: dominated by an old population Does not show enhanced s- process Nor the other pecularities: low Ni, low Na.