Star clusters as age tracers of the age-metallicity relation of the small magellanic cloud

In my Thesis, I determined ages of Small Magellanic Cloud (SMC) star clusters that have formed
during the galaxys entire lifetime. The youngest cluster ages (�10 Myr<age<1 Gyr) were derived using
ground-based photometric data. For the six intermediate-age clusters Lindsay 1, Kron 3, NGC339,
NGC416, Lindsay 38, and NGC419 and the only old globular cluster (GC), NGC121, observations
obtained with the Hubble Space Telescope exists. This work was part of a ground-based and spacebased
program to uncover the age-metallicity evolution of the SMC. In the first three parts of my Thesis,
I presented accurate ages, distance estimates, and structural parameters for the seven intermediateage
and old SMC star clusters. The cluster ages were determined fitting di�erent isochrone models
to the observed color-magnitude diagrams (CMDs). The CMDs reach at least 3 mag below the mainsequence
turno�-points, which makes it the deepest available photometry for these clusters obtained
so far. Only for a few SMC clusters ages had been determined in previous studies using spacebased
data. The ground-based spectroscopy was obtained with the Very Large Telescope (VLT).
My photometric results are combined with these spectroscopic metallicity determinations to obtain a
well-sampled age-metallicity relation. I measured structural parameters of these seven star clusters
and extended the sample of known SMC clusters having accurate age measurements and structural
parameters enormously.
The SMC hosts a large number of intermediate-age and young star clusters, but only one ‘old’
GC, NGC121, for which I determined an age of �10.5 Gyr. Consequently, NGC121 is 2–3 Gyr
younger than the oldest GCs in the Large Magellanic Cloud (LMC) and the Milky Way (MW). For
comparison, the GC system of the MW exhibits a range of ages between �10.5 and 14 Gyr, similar
to the LMC, with the oldest populations belonging to the most ancient surviving stellar systems.
NGC121 is similar in age to the youngest GC in the Fornax dSph and to several of the young Galactic
halo clusters. On the other hand, NGC121 is not as young as some of the Sgr dwarf galaxy’s GCs or
the youngest Galactic GCs. With the SMC having no ‘truly’ old star cluster, it appears that the SMC
has a delayed cluster formation compared to its companions.
The isochrone that fitted best the CMD of NGC121 was �-enhanced. NGC121 is the only known
�-enhanced star cluster in the SMC, a property that it shares with many of the old outer Galactic
halo globulars and which indicates an early rapid chemical enrichment. In a subsequent project
additional SMC clusters will be analyzed for possible �-abundances and elements produced through
r-processes (rapid neutron capture). These �-elements and r-process elements are synthesized in
quickly evolving high-mass stars. The outcoming results can be compared to chemical evolution
models, which calculate chemical abundances in detail (e.g., Pagel & Tautvaiˇsien˙e 1998). With these
models the chemical evolution of the SMC can be further constrained and analyzed.
It is also intriguing that NGC121 is not as metal-poor as the oldest LMC andMWglobulars. The
SMC must have experienced substantial enrichment prior to the formation of NGC121. In the LMC,
two main epochs of the formation of compact populous clusters are observed. In the first epoch, GCs
with ages and metallicities similar to the oldest MW GCs are found, but also very old GCs having a
metallicity similar to NGC121, indicating very early chemical enrichment also in this galaxy. In the second epoch, numerous stellar populations with ages less than 3-4 Gyr developed. The two epochs
are separated by an ‘age gap’ of �4-9 Gyr in which no star cluster has formed. Also theMWcontains
old GCs that have similarly high metallicities as the younger NGC121. Evidently, the conditions
for and the e�ciency of star formation varied in these three galaxies at early epochs. The fact that
this cluster is younger than the Galactic mean, relatively metal-rich, and enhanced in �-elements has
interesting implications for the early development of the SMC.
The SMC is the only dwarf galaxy in the Local Group in which populous star clusters formed and
survived for most of its lifetime. The intermediate-age clusters in the SMC appear to be capable of
surviving a Hubble time, due to their high mass and the structure of the SMC (no bulge or disk to be
passed). After the formation of NGC121, there is a gap of �3 Gyr and thus likely in cluster formation
activity. The second oldest star cluster is Lindsay 1 for which I determined an age of �7.5 Gyr. Since
then compact populous star clusters formed fairly continuously until the present day in the SMC -
a contrast to both the LMC and the MW. For the youngest cluster in my sample, NGC419, me and
Elena Sabbi found indication for a multiple stellar population. A more detailed analysis is in progress
and the study will be published by Elena Sabbi. Only a few multiple stellar populations are known in
theMW, the LMC, and the SMC, but their number is increasing also due to the improved instruments.
Combining the newly derived ages with age and metallicity estimates adopted from di�erent sources
in the literature, it is possible to present a well-sampled age-metallicity relation (AMR) for the SMC,
which is fully based on space-based age determinations and spectroscopic metallicity measurements.
The SMC has experienced an early enrichment as can be seen in the relatively metal-rich oldest
SMC star cluster NGC121. The most striking feature in the AMR is the wide metallicity spread for
clusters with ages around 6 Gyr indicating that the SMC was not very well mixed in the past. The
mean metallicity, however, remains relatively constant for about 4 Gyr, but rises for star clusters that
have formed within the past 2–3 Gyr to a present day metallicity of [Fe/H]�-0.70.
From the apparent magnitudes of the cluster’s red clumps, I provided an estimate of direct distances
for the clusters. Together with cluster distances from the literature that were obtained using
the same approach and that are based on space-based observations, I confirmed the large depth extent
of the SMC along the line-of-sight. The three oldest clusters (age>7 Gyr) are located in the
north-western part of the SMC. NGC361 is a candidate for having an age older than 8 Gyr, but the
age determination found in the literature is associated with large uncertainties and new space-based
photometry of this cluster is needed. The youngest clusters (age<1 Gyr) lie near the SMC main body
in active star forming regions.
The number of intermediate-age and old SMC clusters having accurate structural parameters
and reliable ages was extended enormously in this study. The galactic environment causes external
perturbations such as tidal shocking that occurs as star clusters cross the disk or pass near the bulge.
These processes tend to decrease the cluster mass and therefore change its structural parameters. I
confirmed previous findings (Mackey & Gilmore 2003a,b) that some of the older objects in LMC and
SMC have experienced a significant change in core radius, while for other old objects the core radii
apparently have almost remained unchanged. The core radii of SMC clusters show a trend of older
clusters having a larger spread in core radii than the younger population. Even though I extended
the sample with structural parameters from the literature, the sample is highly incomplete, because
only for a few intermediate-age clusters both reliable ages and corresponding profiles are available.
The analysis of structural parameters of additional SMC clusters is necessary. Clusters in the LMC
have experienced a similar evolution, even though the two galaxies show strong di�erences in various
other aspects. The two confirmed Sagittarius clusters as well as the five Fornax clusters show the same
spread in core radii. The oldest clusters in the MW, however, modified their original structure during
their lifetime and have developed small cores. The largest di�erence between GCs in the MW, the
LMC, and the SMC is that the MW GCs clusters are subject to much larger tidal e�ects. The biggest
dynamical influence on most MW globular halo clusters is the tidal shocking that occurs when they
cross the disk of the MW. Tidal shocking is likely much less e�ective in the LMC and probably even
less so in the SMC. Therefore, the main reason for the smaller core radii of MW globulars is the
di�erent morphology of the three galaxies. The di�erent morphologies might also be the reason for the di�erent flattening distributions of star
clusters in the SMC, the LMC, and theMW. SMC clusters are more flattened than clusters in theMW
and even more flattened than those in the LMC. I found that only NGC121 and Lindsay 38 exhibit a
significant flattening. Galactic GCs modify their original structure and become more spherical with
increasing age, while LMC and SMC clusters maintain their original shape. This might be explained
with the di�erent dynamical influence and therefore the varying strength of the tidal field of the parent
galaxy. The tidal fields of the LMC and SMC might not be strong enough to modify the shape of their
clusters significantly. No relation between cluster age, distance from the SMC center, and ellipticity
was found, but this point needs further analysis because only for a few SMC star clusters reliable
ages, ellipticities and distances are available.
Finally, I provided today’s largest catalog of young SMC star clusters containing ages and luminosities.
The catalog covers an age range between 10 Myr and 1 Gyr. Star clusters are claimed to
be produced through strong shock compressions induced by the collision of their host galaxies which
causes enhanced star formation during close encounters. The most recent model calculations (e.g.,
Bekki & Chiba 2005, Kallivayalil et al. 2006a,b) showed that the SMC, the LMC, and the MW have
only interacted long enough to produce the Magellanic Stream. The models predict the last close encounter
between LMC and SMC around 200 Myr ago due to which enhanced cluster formation can be
expected. The cluster age distribution combining my results with the cluster ages provided by Chiosi
et al. (2006) shows indeed evidence for episodic star formation. The second of two peaks in the age
distribution coincides with the model predicted closest approach of the LMC. The origin of the first
peak about 6.5 Myr ago might have been triggered by internal mechanisms. Looking at their spatial
distribution, the young clusters are assembled in the two large star forming HI super-shells and in the
inter-shell region. Their formation might have been triggered by the expansion of the shells through
gas compression. I found no indication of cluster dissolution. As mentioned above, SMC clusters
evolve di�erently from MW clusters. Due to the di�erent morphologies of the parent galaxies, the
tidal field of the SMC has no big influence on its star clusters. It is most likely that SMC clusters decrease their mass through stellar evolution with time until the clusters finally dissolve.