PaperlayerView exact revision on arXiv

Source-linked AI summary

Chemical Enrichment in the Faintest Galaxies: the Carbon and Iron Abundance Spreads in the Bo\"otes I Dwarf Spheroidal Galaxy and the Segue 1 System

John E. Norris, Rosemary F. G. Wyse, Gerard Gilmore, David Yong, Anna Frebel, Mark I. Wilkinson, V. Belokurov, Daniel B. Zucker

arXiv:1008.0137v1astro-ph.GA

TL;DR

The paper investigates carbon and iron abundance spreads in the faintest Milky Way systems, where Segue 1’s status and candidate membership are uncertain. Using spectroscopic observations of candidate red giants, it finds substantial carbon spreads in Boötes I and Segue 1, alongside iron spreads in both.

  • Problem

    Abundance spreads and the membership of candidate stars in the faintest systems remain uncertain, while Segue 1 has been proposed as either an ultra-faint dwarf or material originating in Sagittarius.

    Source: §1 · p. 3, §3.1.2 · p. 8

  • Method

    The authors use AAOmega spectra of candidate red giants to assess radial-velocity membership and measure carbon and iron abundances in Boötes I and Segue 1.

    Source: §1 · p. 3

  • Results

    Δ[C/H] is ∼0.8 in Boötes I and ∼1.0 in Segue 1, while Δ[Fe/H] is ∼1.6 in both systems.

    Source: §9 · p. 24

  • Takeaways & Limitations

    The luminosity-metallicity relation indicates that the observed dwarf galaxies are not drastically stripped remnants of initially more luminous systems.

    Source: §9 · p. 24

    Segue 1 membership is uncertain: extrapolation suggests up to four of nine candidates could be field contaminants, and no robust ab initio spatial distribution model is available.

    Source: §3.1.2 · p. 9

Abstract

from arXiv · show

We present an AAOmega spectroscopic study of red giant stars in Bootes I, which is an ultra-faint dwarf galaxy, and Segue 1, suggested to be either an extremely low-luminosity dwarf galaxy or a star cluster. Our focus is quantifying the mean abundance and abundance dispersion in iron and carbon, and searching for distant radial-velocity members, in these systems. The primary conclusion of our investigation is that the spread of carbon abundance in both Bootes I and Segue 1 is large. For Bootes I, 4 of our 16 velocity members have [C/H] < ~-3.1, while 2 have [C/H] > ~-2.3, suggesting a range of Delta[C/H] ~ 0.8. For Segue 1 there exists a range Delta[C/H] ~ 1.0, including our discovery of a star with [Fe/H] = -3.5 and [C/Fe] = +2.3, which is a radial velocity member at a distance of 4 half-light radii from the system center. The accompanying ranges in iron abundance are Delta[Fe/H] ~ 1.6 for both Bootes I and Segue 1. For [Fe/H] < -3.0, the Galaxy's dwarf galaxy satellites exhibit a dependence of [C/Fe] on [Fe/H] which is very similar to that observed in its halo populations. We find [C/Fe] ~ 0.3 for stars in the dwarf systems that we believe are the counterpart of the Spite et al. (2005) ``unmixed'' giants of the Galactic halo and for which they report [C/Fe] ~ 0.2, and which presumably represents the natal relative abundance of carbon for material with [Fe/H] = -3.0 to -4.0. We confirm the correlation between luminosity and both mean metallicity and abundance dispersion in the Galaxy's dwarf satellites, which extends to at least as faint as Mv = -5. The very low mean metallicity of Segue 1, and the high carbon dispersion in Bootes I, consistent with inhomogeneous chemical evolution in near zero-abundance gas, suggest these ultra-faint systems could be surviving examples of the very first bound systems.

1. INTRODUCTION

The paper investigates abundance spreads in Boötes I and Segue 1 against a backdrop of questions about the chemistry and nature of the faintest Milky Way satellites. It reports on carbon and iron abundances and considers the implications for these systems’ formation and evolution.

1. INTRODUCTION

  • The nature of the faintest systems remains contested: Segue 1 has been identified as an ultra-faint dwarf galaxy or proposed as a star cluster originating in Sagittarius.Milky Way dwarf spheroidals also offer potential insight into chemical enrichment and the formation of Galactic halo populations.Source: §1 · p. 2
  • Earlier studies found very metal-poor stars and abundance spreads in ultra-faint dwarfs, while subsequent work reported carbon variation and element-abundance patterns resembling those of halo giants.Reported examples include stars with [Fe/H] as low as -3.3 and carbon-enhanced stars in Ursa Major II.Source: §1 · p. 3
  • The study reports a large carbon-abundance range in Boötes I and carbon and iron abundance spreads in Segue 1, including a star comparable to rare carbon-enhanced metal-poor halo stars.The comparison concerns CEMP stars with [Fe/H] ≲ -3.5.Source: §1 · pp. 3–4
  • The paper examines abundance ranges in Boötes I and Segue 1, quantifying carbon abundances for candidate members and considering the implications for ultra-faint galaxy evolution.It also reconsiders membership of two apparently carbon-rich, extremely metal-poor Segue 1 stars.Source: §1 · pp. 3–4

2. OBSERVATIONAL MATERIAL

The study uses AAOmega spectra of candidate red giants in Boötes I and Segue 1. Targets were selected from SDSS data, and the observations sampled stars far from both system centers.

2. OBSERVATIONAL MATERIAL

  • AAOmega blue spectra had resolution R = 5000 and covered 3850–4540 Å; targets were selected from SDSS DR4 using color-magnitude-diagram masks from each system’s discovery paper.The observations used the Anglo-Australian Telescope’s fiber-fed AAOmega spectrograph, which can observe 400 targets over a two-degree field.Source: §2 · p. 4

2.1. Bo¨otes I

  • Boötes I observations yielded useful spectra for stars with 17.5 ≲ g ≲ 21, with the input target list extending to one degree, or four half-light radii, from the galaxy center.Only stars on the red giant locus are discussed and analyzed.Source: §2.1 · p. 4

2.2. Segue 1

  • Segue 1’s target sample contained 323 stars and extended to 40′, or approximately nine half-light radii, from the center; useful spectra covered 17 ≲ g ≲ 21.5.The observed stars spanned a wide color range because the red giant branch location was poorly known.Source: §2.2 · p. 5

3. RADIAL VELOCITIES AND MEMBERSHIP

AAOmega radial velocities identify candidate members in both fields, but velocity errors and foreground complexity limit membership certainty, especially for Segue 1.

3.1.1. Bo¨otes I

  • 36 Boötes I stars fell within 85–130 km s−1, with a mean velocity of 105 km s−1 and a dispersion of 20 km s−1; about four may be Galactic contaminants.The 122-star sample has typical combined internal and external velocity errors of ∼10 km s−1.Source: §3 · p. 6, §3.1.1 · p. 7
  • Six Boötes I candidates lie beyond 35′, but velocity alone provides only moderate membership confidence there.A Gaussian field model predicts four chance matches among 56 distant stars; high-resolution data preliminarily support membership for Boo–980 at 3.9rh.Source: §3.1.1 · pp. 7–8

3.1.2. Segue 1

  • Nine Segue 1 candidates fall between 185 and 230 km s−1, and candidate members extend beyond the nominal half-light radius.The authors’ velocity errors are too large to resolve Segue 1’s internal kinematics.Source: §3 · p. 6, §3.1.2 · p. 8
  • Up to four of the nine Segue 1 candidates may be field contaminants, and no robust joint position–velocity membership criterion is available.The Besançon model is of limited use along this line of sight because of the Sagittarius stream.Source: §3.1.2 · p. 9
  • A luminosity-based estimate predicts four Segue 1 RGB stars, while seven candidates are brighter than the estimate’s limit, suggesting around half may be non-members.The luminosity estimate and membership are uncertain, and the authors cannot identify individual contaminants or resolve the ambiguity with their abundance data.Source: §3.1.2 · pp. 9–10
  • A broad field peak near 300 km s−1 is not predicted by standard Galactic-kinematics models and was tentatively associated with Sagittarius tidal debris.Geha et al. (2009) tentatively identify it with tidal debris from the Sagittarius dSph.Source: §3.1.2 · p. 10

4. STARS WITH RELATIVELY HIGH-S/N AAOMEGA SPECTRA

The section examines higher-S/N spectra of selected velocity candidates in Boötes I and Segue 1. The Segue 1 spectra show substantial differences in carbon- and calcium-sensitive features, with more than one possible explanation for weak calcium lines.

4. STARS WITH RELATIVELY HIGH-S/N AAOMEGA SPECTRA

  • 16 Boötes I stars met the high-S/N and 90–115 km s−1 velocity criteria, with a mean velocity of 105 km s−1 and a dispersion of 6.5 km s−1.Expanding the membership velocity limits would admit two additional objects, but their abundance data are incomplete or uncertain.Source: §4 · p. 10
  • Five Segue 1 candidates met the stated S/N and velocity criteria, and their spectra show a wide range in G-band strength and Ca II H & K lines.Seg 1–7 and Seg 1–98 have weak, ill-defined calcium lines and strong G-bands, resembling carbon-rich extremely metal-poor halo giants.Source: §4 · p. 11
  • Weak Ca II lines in Seg 1–7 and Seg 1–98 could reflect low S/N or emission cores, rather than their abundances alone.The authors found no good spectral match to comparison stars whose weak H & K absorption results from core emission.Source: §4 · p. 11

5. ATMOSPHERIC PARAMETERS

The study estimates stellar atmospheric parameters and metallicities using photometric colors, isochrones, and two abundance methods. Segue 1’s parameter estimates are less consistent than Bootes I’s, and one star is excluded from the candidate giant sample.

Atmospheric parameters

  • The analysis estimates Teff and log g from ugriz colors using synthetic colors and 12-Gyr Yale–Yonsei isochrones, assuming stars lie on the system’s red giant branch.Absolute visual magnitude and the same isochrones provide a check on the color-based estimates.Source: §5 · p. 12
  • For Bootes I, color- and absolute-magnitude-based parameters agree within mean differences of 70 K in Teff and 0.2 in log g; Segue 1’s differences are 230 K and 0.5 before excluding Seg 1–42.Excluding Seg 1–42 reduces Segue 1’s mean differences to 140 K and 0.4.Source: §5 · p. 13

Metal abundances

  • For metallicity, the study uses the Beers et al. (1999) K′ and (B–V)0 calibration or high-resolution model-atmosphere analyses, preferring UVES-based values when available.The calibration assumes the [Ca/Fe]–[Fe/H] relation is shared by Galactic calibration stars and dwarf satellites, which the text says appears to hold below [Fe/H] = –2.0.Source: §5 · p. 13

Metallicity and membership

  • The more-enriched Segue 1 candidates have [Fe/H] values close to typical field-halo stars, complicating membership discrimination at [Fe/H] ~ –1.5 to –2.The text also regards Beers-calibration values for Seg 1–7 and Seg 1–98 as uncertain at this stage.Source: §5 · p. 14

Segue 1 parameters

  • Seg 1–42 is excluded as a red-giant member because its high [Fe/H] = –1.5 conflicts with its unusually high temperature under the assumed giant-branch interpretation.For the four remaining candidate Segue 1 giants, adopted mean errors are σ(Teff) = 140 K and σ(log g) = 0.4.Source: §5 · p. 15

6. FURTHER ASSESSMENT OF THE PUTATIVE SEGUE 1 C-RICH, EXTREMELY METAL-POOR STARS

Follow-up data changed the treatment of the two putative carbon-rich, extremely metal-poor Segue 1 candidates. Seg 1–7 is retained with a preliminary high-resolution abundance estimate, while Seg 1–98 is set aside because of an inconsistency in its metallicity estimates.

Abundance comparisons

  • For seven stars with both AAOmega and VLT/Flames abundances, the mean [Fe/H] difference is –0.06 dex and the dispersion of differences is 0.44 dex.For four stars also measured with Keck/HIRES, the corresponding values are –0.01 and 0.13 dex.Source: §6 · p. 15

Treatment of Seg 1–98

  • Seg 1–98 is excluded after near-infrared spectra examined by the authors yielded substantially higher [Fe/H] than their estimate; further data are needed to resolve the inconsistency.The authors note the rarity of C-rich stars with [Fe/H] ≲ –3.3 and the potential implications of such objects in ultra-faint dwarfs for those systems as building blocks of the outer halo.Source: §6 · pp. 15–16

High-resolution follow-up

  • Seg 1–7’s preliminary UVES analysis gives radial velocity 204.3 ± 0.1 km s−1, [Fe/H] = –3.5, and [C/Fe] ~ +2.5, and the star remains in the analysis.The authors state that the [Fe/H] estimate agrees with the value reported earlier.Source: §6 · p. 16

7. CARBON ABUNDANCES

Carbon abundances are derived by fitting synthetic spectra to the CH G-band, using adopted atmospheric parameters and an assumed oxygen abundance. A comparison with a well-studied metal-poor giant and an independent analysis provides checks on the procedure.

Carbon abundance analysis

  • Synthetic spectra across 4270–4335 Å were fitted to observations of the G-band to determine carbon abundances, using 1D/LTE models and CH oscillator strengths adjusted to fit the Sun.The analysis assumed [O/Fe] = +0.6 because oxygen abundances were unavailable for Bootes I and Segue 1.Source: §7 · p. 16

Validation

  • The procedure gives [C/Fe] = –0.35 for HD 122563, in accord with the reported value of –0.4.The test used the star’s observed spectrum and atmospheric parameters, with adopted [O/Fe] = +0.6.Source: §7 · pp. 16–17

Abundance results and accuracy

  • Estimated [C/H] accuracy is 0.2 dex for Bootes I and 0.4 dex for Segue 1, including atmospheric-parameter propagation and fitting error.For Boo−1137, the table value [C/H] = –3.46 agrees with the independent high-resolution result [C/H] = −3.41.Source: §7 · p. 17

8.1. Carbon Abundances in the Dwarf Spheroidal Galaxies of the Milky Way

The paper reports substantial carbon-abundance variation in Bootes I and Segue 1, and compares their relative carbon abundances with those of the Galactic halo. It discusses evolutionary mixing as an interpretive complication and identifies a similarity between dwarf-galaxy and halo carbon distributions, while noting the need for additional abundance data.

8.1.1. The Spread of [C/H] in Bo¨otes I and Segue 1

  • Bootes I has Δ[C/H] ~ 0.8, with four of 16 stars at [C/H] ≲ –3.1 and two at [C/H] ≳ –2.3; Segue 1 has Δ[C/H] ~ 1.0.The authors say these ranges are relatively robust to representative atmospheric-parameter errors.Source: §8.1.1 · p. 17

8.1.2. Relative Carbon Abundance, [C/Fe], as a Function of Metallicity and Absolute Magnitude

  • The dwarf-galaxy and halo samples show a strong similarity in [C/Fe] behavior, but the halo comparison sample is incomplete above [Fe/H] = –2.5 and is not unbiased.The figure compares [C/Fe] against [Fe/H] and absolute magnitude across dwarf galaxies, halo stars, globular clusters, and ω Cen; the authors caution that the chemically selected ω Cen sample cannot establish a general similarity.Source: §8.1.2 · pp. 18–19, §8.1.4 · p. 20

8.1.3. Red Giants as Probes of Galactic Carbon Enrichment

  • Interpreting red-giant carbon abundances as records of enrichment is complicated by evolutionary mixing, which can lower [C/Fe] as stars ascend the giant branch.The authors point to M15 and M92 as showing declining [C/Fe] alongside increasing [N/Fe], consistent with CN-cycle mixing, while the reason for differences from halo and Bootes I giants remains unclear.Source: §8.1.3 · p. 19

8.1.4. Similarities between the [C/Fe] Distributions of the Ultra-Faint Dwarf Galaxies and the Galactic Halo

  • For –3.5 < [Fe/H] < –2.5, ultra-faint dwarf stars have [C/Fe] ~ 0.3, similar to halo stars; the authors suggest this may correspond to the halo’s “unmixed” giants and natal carbon abundance.Testing that interpretation requires additional nitrogen and lithium measurements in dwarf galaxies.Source: §8.1.4 · p. 20

8.2. Metallicity Distribution Functions

Boötes I has an extended, slowly rising metallicity distribution and may show a radial abundance gradient, while Segue 1’s small sample suggests a broad iron range but leaves its distribution uncertain. Across dwarf galaxies, mean metallicity decreases and abundance dispersion appears to increase toward lower luminosity, though the Segue 1 trend depends on membership assignments.

8.2.1. Bo¨otes I

  • [Fe/H] spans –3.7 to –2.3 in Boötes I, whose MDF rises slowly toward its peak and has the highest fraction of [Fe/H] < –3.0 among dSphs with published individual-star data.The data suggest a radial abundance gradient, but the sample is too small for a strong conclusion.Source: §8.2.1 · p. 21

8.2.2. Segue 1

  • Three objects have reliable abundances and membership-consistent velocities: Seg 1–31 and Seg 1–71 are near [Fe/H] ∼–2.0, while carbon-rich Seg 1–7 is –3.5, making a large iron range probable, not certain.The four velocity-member stars may include a chance non-member, and four objects are too few to characterize the MDF. Whether Segue 1 is an ultra-faint dwarf, tidal debris, or a mixture remains open.Source: §8.2.2 · p. 21

8.2.3. Comparison of metallicity means and dispersions with those of other Milky Way dwarf galaxies

  • The comparison compiles dwarf-satellite and ω Cen means and dispersions using chemically unbiased samples, prioritizing iron as a primary observable and correcting dispersion uncertainties for measurement errors.Adopted measurement errors differ by sample, including 0.15 dex for ten high-resolution Boötes I analyses and 0.35 dex for six AAOmega-based results.Source: §8.2.3 · p. 22
  • Mean [Fe/H] decreases and abundance dispersion appears to increase toward lower luminosity, but whether the trend reaches Segue 1 is uncertain because membership assignments may change it.The authors describe increased spread as intuitively consistent with inhomogeneous stochastic enrichment by very few supernovae.Source: §8.2.3 · p. 23

8.3. The Carbon-Rich, Extremely-Metal-Poor Star in Segue 1

The discussion places carbon-rich, extremely metal-poor stars in the context of Galactic-halo abundance patterns and proposed early supernova enrichment. It notes that such supernovae may also have left signatures in the most metal-poor stars in ultra-faint systems.

8.3. The Carbon-Rich, Extremely-Metal-Poor Star in Segue 1

  • All known Galactic-halo stars with [Fe/H] ≲–4.3 are carbon-rich, and many carbon-enhanced stars below [Fe/H] ≲–3.5 show enhanced C, N, O, or Mg abundances.The increasing carbon-rich fraction toward the lowest halo abundances motivates discussion of their origin (Norris et al. 2007).Source: §8.3 · p. 23
  • Proposed mixing-and-fallback supernovae could account for the enhanced abundance patterns of many carbon-enhanced stars below [Fe/H] ≲–3.5 (Iwamoto et al. 2005).Zero-metallicity massive-star explosions with a Salpeter IMF over 15–40M⊙ do not produce the extreme CNO abundances seen in a significant fraction of these stars (Joggerst et al. 2010).Source: §8.3 · p. 23
  • If more exotic supernovae naturally occur in extremely metal-poor massive stars, they may have left signatures in the lowest-abundance stars of ultra-faint systems.Source: §8.3 · p. 23

9. DISCUSSION AND SUMMARY

Boötes I and Segue 1 show large carbon and iron abundance ranges, while dwarf galaxies exhibit decreasing mean metallicity with decreasing luminosity. The evidence for Segue 1 remains limited by possible contamination and uncertain system membership, but its properties are consistent with an exceptionally low-luminosity, primitive galaxy.

9. DISCUSSION AND SUMMARY

  • ∆[C/H] is ∼0.8 in Boötes I and ∼1.0 in Segue 1, while ∆[Fe/H] is ∼1.6 in both systems.Segue 1 contains a carbon-rich star with [Fe/H] = –3.5 and [C/Fe] = +2.3.Source: §9 · p. 24
  • Dwarf-galaxy mean metallicity decreases with decreasing luminosity, a relation consistent with models involving either mass-dependent ejecta retention or differing gas-loss rates and star-formation histories.Abundance dispersion and abundance-ratio dispersion are needed to assess mixing efficiency and how many supernova ejecta were retained during continued star formation.Source: §9 · p. 24
  • The luminosity–metallicity relation’s continuation below Boötes I remains unproven: available data fit either continuation to Segue 1 at MV = –2 or flattening at Boötes I at MV = –6.Source: §9 · p. 24
  • ω Cen is unique among the displayed metallicity distributions in lacking a tail of very low-metallicity stars, which the authors attribute to formation from pre-enriched gas.They argue this weighs against ω Cen being the remnant core of a typical dSph and favor its interpretation as a luminous star cluster.Source: §9 · p. 25

9.1. What is Segue 1?

  • Segue 1’s four candidate members include two near [Fe/H] ∼−2 and two near [Fe/H] ∼−3.5, but their offsets and possible contamination leave its abundance range uncertain.The two metal-poor stars are offset in velocity or position from reported systemic values.Source: §9.1 · p. 25
  • If at least three of four stars are members, Segue 1’s broad iron range would support self-enrichment in a sufficiently massive dark halo; alternative membership assignments imply a star cluster or a cluster embedded in Sgr debris.Source: §9.1 · pp. 25–26
  • The authors slightly favor Segue 1 as an exceptionally low-luminosity galaxy, while its apparent bimodal abundance distribution and interpretation remain provisional pending larger data sets.They suggest the pattern could reflect a low early star-formation rate or highly inhomogeneous mixing; its very low representative metallicity is consistent with an early system.Source: §9.1 · p. 26
Loading 1008.0137v1…