University of Groningen
The chemical evolution of the dwarf spheroidal galaxy Sextans
Theler, R.; Jablonka, P.; Lucchesi, R.; Lardo, C.; North, P.; Irwin, M.; Battaglia, G.; Hill, V.;
Tolstoy, E.; Venn, K.
Published in:
Astronomy & astrophysics
DOI:
10.1051/0004-6361/201937146
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Final author's version (accepted by publisher, after peer review)
Publication date:
2020
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Theler, R., Jablonka, P., Lucchesi, R., Lardo, C., North, P., Irwin, M., Battaglia, G., Hill, V., Tolstoy, E.,
Venn, K., Helmi, A., Kaufer, A., Primas, F., & Shetrone, M. (2020). The chemical evolution of the dwarf
spheroidal galaxy Sextans.
Astronomy & astrophysics
,
642
, [A176]. https://doi.org/10.1051/0004-
6361/201937146
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.
More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-
amendment.
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum.
Download date: 10-08-2022
Astronomy & Astrophysics manuscript no. 37146corr-PJ
c
ESO 2020
June 11, 2020
The chemical evolution of the dwarf spheroidal galaxy Sextans
???
R. Theler
1
, P. Jablonka
1, 2
, R. Lucchesi
1
, C. Lardo
1
, P. North
1
, M. Irwin
3
, Battaglia G.
4, 5
, V. Hill
6
, E. Tolstoy
7
, K.
Venn
8
, A. Helmi
7
, A. Kaufer
9
, F. Primas
10
, and Shetrone M.
11
1
Physics Institute, Laboratory of Astrophysics, Ecole Polytechnique Fédérale de Lausanne (EPFL), 1290 Sauverny, Switzerland
e-mail: pascale.jablonka@epfl.ch
2
GEPI, Observatoire de Paris, Université PSL, CNRS, Place Jules Janssen, F-92190 Meudon, France
3
Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge CB3 0HA, U.K.
4
Instituto de Astrofísica de Canarias (IAC), Calle Via Láctea, s/n, 38205, San Cristóbal de la Laguna, Tenerife, Spain
5
Departamento de Astrofísica, Universidad de La Laguna, 38206, San Cristóbal de la Laguna, Tenerife, Spain
6
Laboratoire Lagrange, Université de Nice Sophia-Antipolis, Observatoire de la Côte d’Azur, France
7
Kapteyn Astronomical Institute, University of Groningen, Landleven 12, NL-9747AD Groningen, the Netherlands
8
Department of Physics and Astronomy, University of Victoria, PO Box 3055, STN CSC, Victoria BC V8W 3P6, Canada
9
European Southern Observatory, Alonso de Cordova 3107, Vitacura, Casilla 19001, Santiago, Chile
10
European Southern Observatory, Schwarzschild-Str. 2, 85748 Garching, Germany
11
McDonald Observatory, University of Texas at Austin, Fort David, TX, USA
Received XX XX, XXXX; accepted XX XX, XXXX
ABSTRACT
We present our analysis of the FLAMES dataset targeting the central 25
0
region of the Sextans dwarf spheroidal galaxy (dSph). This
dataset is the third major part of the high-resolution spectroscopic section of the ESO large program 171.B-0588(A) obtained by the
Dwarf galaxy Abundances and Radial-velocities Team (DART). Our sample is composed of red giant branch stars down to V∼20.5
mag, the level of the horizontal branch in Sextans, and allows users to address questions related to both stellar nucleosynthesis and
galaxy evolution.
We provide metallicities for 81 stars, which cover the wide [Fe/H]=−3.2 to −1.5 dex range. The abundances of ten other elements
are derived: Mg, Ca, Ti, Sc, Cr, Mn, Co, Ni, Ba, and Eu. Despite its small mass, Sextans is a chemically evolved system, showing
evidence of a contribution from core-collapse and Type Ia supernovae as well as low-metallicity asymptotic giant branch stars (AGBs).
This new FLAMES sample offers a sufficiently large number of stars with chemical abundances derived with high accuracy to firmly
establish the existence of a plateau in [α/Fe] at ∼ 0.4 dex followed by a decrease above [Fe/H]∼ −2 dex. These features reveal a close
similarity with the Fornax and Sculptor dSphs despite their very different masses and star formation histories, suggesting that these
three galaxies had very similar star formation efficiencies in their early formation phases, probably driven by the early accretion of
smaller galactic fragments, until the UV-background heating impacted them in different ways. The parallel between the Sculptor and
Sextans dSph is also striking when considering Ba and Eu. The same chemical trends can be seen in the metallicity region common
to both galaxies, implying similar fractions of SNeIa and low-metallicity AGBs. Finally, as to the iron-peak elements, the decline
of [Co/Fe] and [Ni/Fe] above [Fe/H]∼ −2 implies that the production yields of Ni and Co in SNeIa are lower than that of Fe. The
decrease in [Ni/Fe] favours models of SNeIa based on the explosion of double-degenerate sub-Chandrasekhar mass white dwarfs.
Key words. stars: abundances / galaxies: individual: Sextans dwarf spheroidal / galaxies: evolution
1. Introduction
A large variety of topics rely on our understanding of the for-
mation and evolution of dwarf galaxies. In the ΛCDM cosmol-
ogy, galaxies form hierarchically by cooling and condensation
of the baryons within dark-matter haloes that gradually merge
(White & Rees 1978). Strong efforts are therefore concentrated
on counting, characterising, and quantifying the impact of the
earliest and smallest of these galactic systems from the local uni-
verse to the reionisation period (e.g. Tolstoy et al. 2009; Wise
et al. 2014; Sawala et al. 2016; Simon 2019; Torrealba et al.
2019).
The Sextans dwarf spheroidal galaxy (dSph) was discovered
by Irwin et al. (1990). At a distance of ∼ 90 kpc, it is one of the
closest satellites of the Milky Way (Mateo et al. 1995; Lee et al.
?
Based on the ESO Program 171.B-0588(A)
??
Tables 2-6 and 9-13 are available in electronic form at the CDS via
anonymous ftp to cdsarc.u-strasbg.fr
2003; Battaglia et al. 2011). Its late discovery is the consequence
of its large extent on the sky with a tidal radius of 120 ± 20
arcmin (Cicuéndez et al. 2018) and low surface brightness of
σ
0
= 18.2 ± 0.5 mag/arcmin
2
(Irwin & Hatzidimitriou 1995)
making it a challenging galaxy to characterise given the large
fraction of Milky Way interlopers.
The analysis of the colour magnitude diagram (CMD) of
Sextans reveals a stellar population which is largely dominated
by stars older than ∼11Gyr (Lee et al. 2009), with evidence for
radial metallicity and age gradients, the oldest stars forming the
most spatially extended component (Lee et al. 2003; Battaglia
et al. 2011; Okamoto et al. 2017; Cicuéndez et al. 2018).
Spectroscopic follow-up in Sextans started in 1991 at
medium–low resolution in the region of the calcium triplet (CaT,
8498, 8542 and 8662 Å). Da Costa et al. (1991) identified
six galaxy members and derived a mean metallicity of [Fe/H]
= −1.7 ± 0.25 dex. Suntzeff et al. (1993) increased the sample of
Article number, page 1 of 36
arXiv:1911.08627v2 [astro-ph.GA] 9 Jun 2020
A&A proofs: manuscript no. 37146corr-PJ
galaxy members up to 43 and revised the galaxy peak metallic-
ity to −2.05 ± 0.04 dex. The enhanced multiplexing power of a
new generation of spectrographs with over ∼ 30
0
fields of view
opened up the possibility to analyse hundreds of stars at once.
Spectroscopy in the region of the Mg I triplet (5140-5180 Å) and
the CaT absorption features were used for rough chemical tag-
ging and to investigate the mass profile of Sextans, as well as the
possible existence of kinematically distinct stellar populations at
its centre (Walker et al. 2006, 2009; Battaglia et al. 2011). Sex-
tans is thought to be about 0.4 Gyr away from its pericentre (r
peri
∼ 75 kpc) moving towards its apocentre (r
apo
∼ 132 kpc), and
its orbit seems inconsistent with a membership to the vast po-
lar structure of Galactic satellites (Casetti-Dinescu et al. 2018;
Fritz et al. 2018). Until recently, no statistically significant dis-
tortions or signs of tidal disturbances had been found down to
very low surface brightness limit (Cicuéndez et al. 2018), but
combined reanalysis of the spectroscopic membership and deep
photometry have revealed the presence of a ring-like structure
that is interpreted as the possible sign of a merger (Cicuéndez &
Battaglia 2018).
One thread of studies traces the formation and evolution
of galaxies by exploring their chemical evolution as preserved
in stellar abundance patterns. Comparison between galaxies of
very different star formation histories also offers important in-
sight into poorly understood nucleosynthetic origins such as for
the neutron capture elements (e.g. Tolstoy et al. 2009; Jablonka
et al. 2015; Mashonkina et al. 2017; Ji et al. 2019; Reichert et al.
2020). Here, spectroscopic multiplex again plays a fundamental
role allowing to switch from the pioneer ensembles of a few stars
per galaxy that had elemental abundances and abundance ratios
(e.g. Hill et al. 1995; Shetrone et al. 2001, 2003) to statistically
significant samples. A step forward arose from Keck/DEIMOS
medium-resolution (R∼7000) spectroscopy, with about 35 Sex-
tans stars with delivered abundances of α-elements (Mg, Si, Ca,
Ti) at accuracies better than 0.3 dex (Kirby et al. 2011). This
sample has recently been completed with Cr, Co, and Ni abun-
dances (Kirby et al. 2018). The number of remaining open ques-
tions in Sextans is nevertheless very large. In particular, its low
metallicity range ([Fe/H]≤ −2.) is still uncovered. Only a few ex-
tremely metal-poor stars have been targeted in this galaxy (Aoki
et al. 2009; Tafelmeyer et al. 2010). The mean trend and scatter
at fixed [Fe/H] of the abundance ratios still need to be investi-
gated over the full chemical evolution of Sextans.
The VLT/FLAMES fibre-spectrograph has already been
transformative in addressing similar questions in other dSphs.
The Dwarf Abundances and Radial velocity team (DART) tar-
geted the Fornax and Sculptor dSphs (Letarte et al. 2010;
Hill et al. 2019). The present paper presents the DART high-
resolution spectroscopic sample of Sextans. In the following, we
present the analysis of a sample of 81 Sextans stars that have
been observed in the central region of Sextans at high resolution
(R∼20,000) with VLT/FLAMES (GIRAFFE and UVES). This is
the largest sample of red giant branch (RGB) stars dedicated to a
chemical analysis, and allows us to address questions related to
both stellar nucleosynthesis and galaxy evolution.
2. Observations and data reduction
Our targets are RGBs located in the central 25
0
field of the Sex-
tans dSph. About half of the sample is composed of RGBs previ-
ously identified as Sextans members based on their radial veloc-
ities measured from medium-resolution spectra around the cal-
cium triplet (Battaglia et al. 2011). The rest of the sample was
selected from the position of the stars in the CMD of Sextans.
−2−1012
δ RA [deg]
−2
−1
0
1
2
δ DEC [deg]
Fig. 1: Spatial distribution of the DART spectroscopic observa-
tions in Sextans. The grey points show the stars observed around
the CaT survey (Battaglia et al. 2011). The blue circles corre-
spond to the probable members of this medium-resolution sam-
ple, while the green points are the stars of our high-resolution
sample (see Table 2 for the values of RA and DEC). The black
ellipse represents the tidal radius of Sextans calculated by Ci-
cuéndez et al. (2018).
The spatial distribution of our sample stars is presented in Fig.
1, while Fig. 2 displays their location along the RGB V versus
V − I plane.
We gathered high-resolution (R∼20’000) spectra of 101 stars
in the HR10, HR13, and H14 gratings of the multi-fibre spectro-
graph FLAMES/GIRAFFE installed at the VLT (ESO Program
171.B-0588(A)). The observations were conducted in three runs
from March to December 2004 and led to a total exposure time
of 30 hours and 17 minutes. Two fibres were linked to the red
arm of the UVES spectrograph, yielding R ∼ 47 000 spectra of
two stars, S05-5 and S08-229, over the λ ∼ 4800-6800 Å range.
Table 1 summarises the characteristics of the gratings and the
corresponding exposure times.
In each MEDUSA plate, 16 (in HR10) and 19 (in HR13,
HR14) fibres were dedicated to the sky background. Five fibres
were allocated to simultaneous wavelength calibration in HR10
and HR13, but not in HR14 to prevent pollution from the argon
lines which saturate at wavelength longer than 6500 Å. Never-
theless, in some cases these wavelength calibration spectra have
also polluted their closest neighbour stellar spectra in HR10 and
HR13. The spectra of 25 stars are affected by this problem; they
are indicated in Table 2. For those stars, we discarded the pol-
luted spectra (HR10 and/or HR13) from further abundance anal-
ysis. Three more stars have one or two missing parts of their
spectra: S05-70 and S05-78 have no HR10 spectrum and S08-
274 has only a HR10 spectrum. This information is also reported
in Table 2.
The data reduction has been performed with the ESO GI-
RAFFE Pipeline version 2.8.9. For each science frame, we used
the corresponding bias, dark, and flat-field frames, as well as
arc lamp spectrum. For the sky subtraction, we used the routine
of M. Irwin (priv. comm.), which creates an average sky spec-
Article number, page 2 of 36
R. Theler et al.: The chemical evolution of the dwarf spheroidal galaxy Sextans
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
V-I
17
18
19
20
21
22
V[mag]
Fig. 2: Colour-magnitude diagram of the Sextans stars. The blue
and green points are in as in Fig. 1. The black dots correspond
to Sextans photometry taken with the CFH12K CCD camera at
Canada-France-Hawaii Telescope (CFHT; Lee et al. 2003).
Grating HR10 HR13 HR14
∆λ ∼5340-5620 Å ∼6120-6400 Å ∼6300-6700 Å
Resolution 19800 22500 17740
Exposure Time 36080 s 43566 s 29400 s
Table 1: Summary of the observations: The three gratings of the
GIRAFFE observations are indicated as well as their wavelength
coverage (∆λ) and total exposure times. Two fibres were simul-
taneously connected for the total exposure time to the red arm of
UVES with a central wavelength of ∼5800 Å and a resolution of
47000.
trum from the sky fibres. This average sky spectrum was then
subtracted from each target spectrum after appropriate scaling in
order to match the sky features in each fibre. We extracted the
individual spectra with the task scopy in IRAF. The heliocentric
velocities were calculated and possible shifts were corrected be-
fore the individual subexposures were averaged with scombine
in IRAF with an averaged sigma clipping algorithm that removes
remaining cosmic rays and bad pixels.
The signal-to-noise ratio (S/N) per pixel of each of our 101
spectra was estimated with the IRAF task splot in three contin-
uum regions: [5456.6Å, 5458.94Å], [6193.47Å, 6197.74Å], and
[6520.0Å, 6524.4Å] for HR10, HR13, and HR14, respectively.
The results are presented in Table 3.
The spectra were normalised with the routine of M. Irwin
(priv. comm.), which consists in an iterative process of spec-
trum filtering. The routine detects the lines through asymmetric
k-sigma clipping and they are masked to fit the continuum.
0 50 100 150 200 250
v
rad
[km/s]
0
2
4
6
8
10
12
14
16
N
Fig. 3: Radial velocity distribution of our stars (black solid line).
The dashed red line corresponds to the mean radial velocity
from Battaglia et al. (2011) based on 174 probable members:
v
sys, helio
= 226.0 ± 0.6 km/s. The two solid red lines represent
the ±3σ interval within which the membership zone is defined.
3. Selection of the Sextans members from the
GIRAFFE sample
We derived the stellar radial velocities from the 1D reduced
spectra with DAOSPEC
1
(Stetson & Pancino 2008), which cross-
correlates all the detected lines with an input line list that we took
as in Tafelmeyer et al. (2010). The radial velocities in each grat-
ing (HR10, HR13, HR14) and their mean are provided in Table
4 and shown in Figure 3.
The Sextans members were selected by requiring that their
radial velocities fall within 3σ of the galaxy systemic velocity
derived by Battaglia et al. (2011), 226.0±0.6 km/s, σ = 8.4±0.4
km/s. Twelve stars were identified as background or foreground
interlopers; they are indicated as Non-member (v) in Table 2.
Our final cleaned sample has a mean velocity and dispersion of
v
sys,helio
= 224.9 ± 1.9 km/s, σ = 7.7 ± 0.5 km/s in perfect agree-
ment with Battaglia et al. (2011).
The dataset has been acquired over a time period of 1 month,
and this time delay between the HR10 and HR13/HR14 spectra
allowed us to look for binaries. These were identified based on
radial velocities derived in each grism differing by more than
1σ, the error on the velocity. This corresponded to differences
in velocities of ∆v
rad,HR
between 7 and 35 km/s. A total of 13
binary systems were identified in this way, for which the mean
radial velocities in Table 2 are preceded by a tilde symbol. We
also identified five carbon stars with a strong C
2
band head. A
preliminary analysis shows that these are CEMP-s stars. Further
investigation of these stars is postponed to another paper.
Five stars (S05-72, S08-321, S08-111, S08-301, and S08-
293) with overly poor-quality spectra and very unstable solutions
for their abundances have also been removed from our investi-
gation. These latter stars usually fall at the margins of the photo-
metric RGB and are classified as Non-member (c) in Table 2. As
a result of the above selection, our final sample encompasses 81
RGB Sextans stars.
1
DAOSPEC has been written by P.B. Stetson for the Dominion As-
trophysical Observatory of the Herzberg Institute of Astrophysics, Na-
tional Research Council, Canada.
Article number, page 3 of 36
A&A proofs: manuscript no. 37146corr-PJ
4. Stellar atmosphere models
Our full sample is distributed along the Sextans RGB down to the
horizontal branch level, that is, reaching uncommonly faint mag-
nitudes for high-resolution spectroscopic analyses. For that rea-
son, the elemental abundances of the GIRAFFE sample were de-
rived by synthesis based on a χ
2
minimisation procedure, which
can balance the derivation of the chemical abundances over the
widest possible wavelength range at once. This procedure is de-
scribed in Sect. 7. Two stars were sufficiently bright to be fed
to UVES. These have been analysed in a classical way, which is
detailed in Sect. 6, based on the equivalent widths of their indi-
vidual absorption lines.
In both cases, we adopted the MARCS 1D spherical atmo-
sphere models, which were downloaded from the MARCS web
site
2
(Gustafsson et al. 2008), and interpolated using the code of
Thomas Masseron available on the same website. We assumed
standard values of [α/Fe] following the Galactic disc and halo,
namely +0.4 for [Fe/H]≤ −1.5, +0.3 for [Fe/H]=−0.75, +0.2 for
[Fe/H]=−0.54, +0.1 for [Fe/H]=−0.25, and +0.0 for [Fe/H] ≥
0.0.
5. Photometric parameters
The photometric estimates of the stellar parameters T
eff
, log g,
and v
turb
were based on the ESO 2.2m WFI V and I-band magni-
tudes (Tolstoy et al. 2004; Battaglia et al. 2006, 2011), as well as
on the J, H, K s WFCAM LAS UKIRT photometry calibrated
onto the 2MASS photometric system (Skrutskie et al. 2006).
These magnitudes are provided in Table 2.
For each star, the photometric effective temperature is taken
as the simple average of the four colour temperatures T
V−I
,
T
V−J
, T
V−H
, and T
V−K
obtained with the calibration of Ramírez
& Meléndez (2005). When possible, the metallicity estimate
from the CaT was used as initial iron abundance; otherwise
we adopted the mean metallicity of the Sextans stellar popu-
lation, [Fe/H]=−1.9, as given by Battaglia et al. (2011). We
used E
B−V
= 0.0477 (Schlegel et al. 1998) and the reddening
law, A
V
= 3.24 E
B−V
, of Cardelli et al. (1989). The stellar sur-
face gravities were estimated from the stellar effective tempera-
ture and the bolometric correction calculated as in Alonso et al.
(1999). We adopted a distance of 90 kpc (Tafelmeyer et al. 2010;
Battaglia et al. 2011), a 0.8 M
stellar mass for our sample stars
and log g
= 4.44, T
eff
= 5790 K, and M
Bol,
= 4.75. The
micro-turbulence velocities were derived according to the em-
pirical relation v
turb
= 2.0 − 0.2 × log g of Anthony-Twarog et al.
(2013). The effective temperature for each colour, the bolometric
corrections and initial metallicities are provided in Table 3.
These photometric parameters were the final ones for the GI-
RAFFE sample, while T
eff
and v
turb
could be further spectro-
scopically adjusted for the two stars observed with UVES as de-
scribed in Sect. 6.
6. Analysis of the two stars observed with UVES
S05-5 and S08-229 were sufficiently bright to be allocated to
two of the eight UVES fibres of the FLAMES configuration.
These stars were analysed in the same way as in our previous
publications on high-resolution spectroscopic studies in dwarf
spheroidal galaxies (e.g. Letarte et al. 2006, 2010; Tafelmeyer
et al. 2010; Jablonka et al. 2015; Hill et al. 2019). The abundance
analysis has been performed using the local thermodynamical
2
marcs.astro.uu.se
equilibrium (LTE) code CALRAI first developed by Spite (1967)
(see also Cayrel et al. (1991) for the atomic part), and contin-
uously updated over the years. We summarise the main steps
below.
6.1. Measurements
The absorption features were measured following the line list
of Tafelmeyer et al. (2010) which combines those of Letarte
et al. (2010) and Cayrel et al. (2004). The equivalent widths
were measured with DAOSPEC. As DAOSPEC fits absorption
lines with Gaussians that have a fixed FWHM, the equivalent
widths of strong lines with prominent wings are systematically
underestimated. Therefore, lines with equivalent widths larger
than 180 mÅ were discarded. Table 5 lists the lines and their
equivalent widths. The present analysis focuses on the elements
derived in the HR10, HR13, and HR14 grisms. A discussion of
the other elements accessible thanks to the UVES wavelength
coverage and high resolution is deffered to a forthcoming paper.
6.2. Final atmospheric parameters and error estimates
The convergence to our final effective temperatures and the mi-
croturbulence velocities (v
turb
) presented in Table 6 was achieved
iteratively, as a trade off between minimising the trends of metal-
licity derived from the Fe i lines with excitation potentials and
equivalent widths on the one hand and minimising the differ-
ence between photometric and spectroscopic temperatures on the
other hand. Starting from the initial photometric parameters we
adjusted T
eff
and v
turb
allowing for deviation by no more than 2
σ (the uncertainty) of the slopes. This yielded new metallicities
which were then fed back into the photometric calibration to get
new photometric temperatures and gravities. No more than two
or three iterations were needed to converge to our final atmo-
spheric parameters.
The predicted equivalent widths rather than the observed
ones were considered in this procedure, because the errors on
the measurements can bias the slope of the diagnostic plots (see
Magain 1984). However, the surface gravities were kept fixed to
their photometric values due to the small number of Fe ii lines
and possible NLTE effects, which might affect the ionisation
equilibrium.
The final abundances are calculated as the weighted mean
of the abundances obtained from the individual lines, where the
weights are the inverse variances of the single line abundances.
These variances were propagated by CALRAI from the estimated
errors on the corresponding equivalent widths. They are listed in
Table 7. Table 8 provides the errors on the abundances linked
to the uncertainties on atmospheric parameters for the two stars
observed with UVES.
7. Analysis of the GIRAFFE sample
7.1. Synthetic grid
[Fe/H]: To determine [Fe/H], a library of synthetic spectra
covering the HR10, HR13, and HR14 wavelength regions was
created with MOOG
3
(August 2010 Version). The stellar spectra
3
MOOG (Sneden 1973) is a FORTRAN code designed to perform a va-
riety of LTE line analyses and spectrum synthesis with the aim being
to help with determination of the stellar chemical composition. The
basic equations of the stellar line analysis in LTE are used follow-
ing the formulation of Edmonds (1969). Much of the MOOG code fol-
Article number, page 4 of 36
![Fig. 13: Comparison between the three classical dwarfs, Sextans (green), Sculptor (blue), and Fornax (red), and the Milky Way (grey) in the region of the knee in [α/Fe].](/figures/fig-13-comparison-between-the-three-classical-dwarfs-sextans-11ugzw1y.webp)
![Table 8: Errors due to uncertainties in atmospheric parameters for S05-5, S08-229 observed with UVES. They are related to [Fe/H] for iron and to [X/Fe] for the other elements.](/figures/table-8-errors-due-to-uncertainties-in-atmospheric-21jw2lrt.webp)
