The Star Formation History of the Sagittarius Dwarf Galaxy and Streams

The Star Formation History of the Sagittarius dwarf galaxy and streams

Thomas de Boer ! V. Belokurov, S. Koposov, N. W. Evans, D. Erkal and many more ! Institute of Astronomy Cambridge - United Kingdom

1 The Sagittarius stream(s)

Sgr is a large and luminous dwarf 9 ->Progenitor mass: ~10 M⊙ (SMC-like) 8 ->Luminosity: ~ 10 L⊙ MV~-15.2 -> 70% of luminosity in stream ! Sgr stream: ->Largest stream in MW halo ->At least 1 full wrap around MW! ! SDSS MSTO stars (Belokurov et al. 2006) Important for studying halo formation through massive systems (and in comparison to LG dSph) !

2MASS M-giants2 (Majewski et al. 2003) One Sgr stream… or two?

Multiple sequences in Sgr stream! ‘bifurcation’ in North? Stream split? ! Sgr stream can be separated in 2 components ->faint stream: diff distance, simpler populations ! Open questions: ->stellar population differences? ->drawn from same progenitor? ->different pericentre passage?

Law & Majewski Sgr stream coordinates

Need to study the stellar content of Sgr!

(Koposov et al. 2012) 3 Photometric stream samples

14 100 Sgr bright stripe82 full ! 15 SDSS Stripe 82 photometry 16 80 17 60 -> single epoch and deep co-add-> photometric 18 i N 19 completeness 40 20 -> Sgr based on �,B selection (Law & Majewski model) 21 20 22 -> MW foreground correction using Galactic-mirrored 23 0 -0.5 0 0.5 1 1.5 2 g-i ﬁelds (same l, inverse b) 14 100 Sgr bright stripe82 MW region -> Distance gradient correction using distances from 15 16 80

Koposov et al. 2012 17 60 18 i ! N 19 40 20 ! 15 a b 21 20 4 22

10 23 0 -0.5 0 0.5 1 1.5 2 2 g-i 14 100 5 Sgr bright stripe82 MW corrected 15 0 80 (deg) 16 ⊙ ~ dec (deg) 0 17 60 18 i

-2 N 19 -5 40 20 -4 21 20 -10 22 60 50 40 30 20 10 0 230 240 250 260 270 280 290 23 0 ~ -0.5 0 0.5 1 1.5 2 ra (deg) (deg) 4 ⊙ g-i Spectroscopic stream samples

Spectroscopic sample from SDSS/SEGUE -> atmospheric parameters (log g, log Teff) -> radial velocities -> metallicity [Fe/H] -> average α-element abundance [α/Fe]

0.12 ! normalised MDF bright gl faint Select Sgr based on:0.1 0.08

-> spatial location0.06 Count -> radial velocity 0.04 gb

0.02 200 -> distance a) 0 -2.9 -2.5 -2.1 -1.7 -1.3 -0.9 -0.5 -0.1 150 -> select only giants (log g<3)[Fe/H] (dex) 100 0.12 normalised MDF bright 1 cumulative MDF faint 0.1 50 0.8

0.08 (km/s) 0 0.6 0.06 GSR V -50 Count Count 0.4 0.04 -100 0.02 0.2 -150 0 0 -2.9 -2.5 -2.1 -1.7 -1.3 -0.9 -0.5 -0.1 -2.9 -2.5 -2.1 -1.7 -1.3 -0.9 -0.5 -0.1 -200 [Fe/H] (dex) [Fe/H] (dex) 100 150 200 250 300 (deg) 1 cumulative MDF 5 ⊙

0.8

0.6 Count 0.4

0.2

0 -2.9 -2.5 -2.1 -1.7 -1.3 -0.9 -0.5 -0.1 [Fe/H] (dex) Bright and faint streams

Combination of spectroscopy and photometry shows clear stellar population picture ! ! MSTO: extended distribution: multiple populations faint stream shows simpler CMD -> simpler stellar populations ! RGB: Bright stream bi-modal extended MDF Faint stream more metal-poor ->lacks strong metal-rich ([Fe/H]>-0.9) component

6 Combining all pieces: the SFH

Combine photometry and spectroscopy directly to constrain ages ! ! Construct synthetic CMD’s -> arbitrary age, [Fe/H], [α/Fe] -> different isochrone sets -> photometric completeness ! Construct synthetic MDFs -> extract stars with similar magnitude range -> bin in [Fe/H] -> convolve with Gaussian ! ! SFH using MSTO photometry (age sensitive) and RGB MDF (direct metallicity) (de Boer et al 2012)

7 Fitting the SFH

! ! Fit single-epoch as well as deep co-add Fit with and without spectroscopy ! Sensible residuals, models reproduce CMD ->overall small residuals (<3 sigma in most bins) ->blue stragglers (g-i<0) ﬁt as young population ->small amount of positive residuals MW subtraction not perfect? ! Solutions without MDF prefer more metal-poor SFH

0.15 0.15 bright single epoch obs faint single epoch obs MDF fit MDF fit 0.12 no MDF 0.12 no MDF

0.09 0.09

0.06 0.06 Normalised count Normalised count 0.03 0.03

0 0 -2.7 -2.3 -1.9 -1.5 -1.1 -0.7 -0.3 0.1 -2.7 -2.3 -1.9 -1.5 -1.1 -0.7 -0.3 0.1 [Fe/H] (dex) [Fe/H] (dex) 8 0.15 0.15 bright coadd obs faint coadd obs MDF fit MDF fit 0.12 no MDF 0.12 no MDF

0.09 0.09

0.06 0.06 Normalised count Normalised count 0.03 0.03

0 0 -2.7 -2.3 -1.9 -1.5 -1.1 -0.7 -0.3 0.1 -2.7 -2.3 -1.9 -1.5 -1.1 -0.7 -0.3 0.1 [Fe/H] (dex) [Fe/H] (dex) SFH of bright Sgr stream

! SFH shows tight sequence in age-[FeH] plane ->stars formed in well-mixed, homogeneously enriched medium. ! Similar results single-epoch and co-add photometry -> MDF adds meaningful constraints on SFH ! Sequence consistent with age and metallicity of GCs associated to Sgr

8 2.5 -> stream stars drawn from same population mix as Sgr bright single epoch no MDF bright single epoch no MDF 7 MDF fit MDF fit 2 ! 6 5 1.5 Change of slope at age 11-13 Gyr, consistent 4 ) ) 2 2 3 1 with Sgr alpha-element knee (de Boer et al. 2014) 2

/yr/deg 0.5

1 /dex/deg ->supernovae Ia started contributing to abundance pattern sun sun

M 0 0 M

-6 1-3 Gyr after start of star formation. bright coadd no MDF 4 bright coadd no MDF 7 MDF fit MDF fit 2 6 ! SFR (10 5 SFR (10 1.5 4 Star formation rate drops sharply at 5-7 Gyr 3 1 -> related to infall of Sgr into the MW? 2 0.5 1 0 0 0 2 4 6 8 10 12 14 -2.7 -2.3 -1.9 -1.5 -1.1 -0.7 -0.3 0.1 Age (Gyr) [Fe/H] (dex) 9 SFH of faint Sgr stream

! Same tight sequence as in bright stream -> Sgr dwarf is progenitor of the faint component as well as the bright one ! Lower S/N of the stream results in the presence of more anomalous populations ->metal-rich populations likely ﬁt to red MW stars ! Faint stream composed of simpler population mix than the bright stream

7 1.6 -> consistent with CMD morphology faint single epoch no MDF faint single epoch no MDF 6 MDF fit 1.4 MDF fit ! 5 1.2 1 4 Sequence dominated by old (>8 Gyr) metal 0.8 ) ) 2

2 3 0.6 poor stars 2 0.4 /yr/deg 1 /dex/deg 0.2 ->stream drawn from more pristine Sgr population mix sun sun

M 0 0 M -6 faint coadd no MDF 4 faint coadd no MDF ->stripped earlier? from the outskirts? 6 MDF fit 1.4 MDF fit 5 1.2 ! SFR (10 SFR (10 1 4 0.8 Earlier pericentre passage of the stream? 3 0.6 2 0.4 1 0.2 0 0 0 2 4 6 8 10 12 14 -2.7 -2.3 -1.9 -1.5 -1.1 -0.7 -0.3 0.1 Age (Gyr) [Fe/H] (dex) 10 Conclusions

First detailed quantitative study of the Sgr trailing stream ! Sgr SFH of both components show a tight sequence in the plane of Age vs [Fe/H] ->star-formation and enrichment proceeded in a similar fashion for each part of the bifurcation. ->star-formation within Sgr took place in a well-mixed medium, homogeneously enriched in metals over 8 Gyr. ! Comparison to Sgr GCs: ->both streams are consistent with Sgr populations ->Sgr dwarf is progenitor of the faint component as well as the bright one ! Star formation rate drops rapidly around 5-7 Gyr ago ->could be caused by the infall of Sgr into the MW, coinciding with stripping of gas ! Faint stream composed of simpler stellar population mix than the bright stream -> dominated by old metal poor stars -> lacking strong metal-rich component found in the bright stream MDF. ! Faint stream likely produced by material stripped earlier and from the outskirts of Sgr.