Detecting metal-rich intermediate-age globular clusters in NGC4570 using K-band photometry
ABSTRACT Globular cluster systems (GCSs) of most early-type galaxies feature two peaks in their optical colour distributions. Blue-peak
globular clusters (GCs) are believed to be old and metal-poor, whereas the ages, metallicities, and the origin of the red-peak
GCs are still being debated. We obtained deep K-band photometry and combined it with Hubble Space Telescope observations in g and z to yield a full spectral energy distribution from the optical to the near-infrared. This now allows us to break the age–metallicity
degeneracy. We used our evolutionary synthesis models galev for star clusters to compute a large grid of models with different metallicities and a wide range of ages. Comparing these
models to our observations revealed a large population of intermediate-age (1–3Gyr) and metal-rich (≈solar-metallicity) GCs,
that will give us further insights into the formation history of this galaxy.
KeywordsGalaxies: elliptical and lenticular, cD-Galaxies: formation-Galaxies: individual: NGC 4570-Galaxies: star clusters
arXiv:0801.2318v1 [astro-ph] 15 Jan 2008
Astrophysics and Space Science
Detecting metal-rich intermediate-age Globular Clusters
in NGC 4570 using K-band photometry
R. Kotulla1•U. Fritze1•P. Anders2
c ? Springer-Verlag ••••
Abstract Globular Cluster Systems (GCSs) of most
early-type galaxies feature two peaks in their optical
colour distributions. Blue-peak GCs are believed to be
old and metal-poor, whereas the ages, metallicities, and
the origin of the red-peak GCs are still being debated.
We obtained deep K-band photometry and combined it
with HST observations in g and z to yield a full SED
from optical to near-infrared. This now allows us to
break the age-metallicity degeneracy.
We used our evolutionary synthesis models GALEV
for star clusters to compute a large grid of models with
different metallicities and a wide range of ages. Com-
paring these models to our observations revealed a large
population of intermediate-age (1–3 Gyr) and metal-
rich (≈ solar metallicity) globular clusters, that will
give us further insights into the formation history of
Globular Cluster Systems (GCSs) are now recog-
nized as powerful tracers of their parent galaxy’s for-
mation history (West et al. 2004; Fritze-v. A. 2004;
Brodie and Strader 2006). From their age and metallic-
ity distributions one can reconstruct the parent galaxy’s
(violent) star formation and chemical enrichment his-
tory all the way from the very onset of star formation
in the early universe to the present.
Most early-type galaxies show a bimodal color distri-
bution for their GCSs (e.g. Gebhardt and Kissler-Patig
1Centre for Astrophysics Research, University of Hertfordshire,
College Lane, Hatfield AL10 9AB, UK
2Sterrekundig Intituut, Universiteit Utrecht, Princetonplein 5,
3584 CC Utrecht, NL
1999; Kundu and Whitmore 2001,?; Peng et al. 2006):
A universal blue peak and a red peak with colors and
height relative to the blue peak varying from galaxy to
galaxy. The blue peak GCs are believed to be old and
metal-poor, the origin of the red peak is still unclear.
1.1 Our approach to lift the age-metallicity
Optical data alone do not allow to disentangle ages
and metallicities:Colour-to-metallicity transforma-
tions have to assume an age, while colour-to-age trans-
formations are only valid for one metallicity. This de-
generacy, however, can be broken by including near-
infrared data that are more sensitive to changes of
metallicity rather than age.
Anders et al. (2004) used extensive artificial star
cluster tests and showed that observations in
• 3 passbands for GCSs in dust-free E/S0s or
• 4 passbands for (young) star clusters in dusty envi-
• spanning as wide as possible a wavelength-basis (U
through K) and
• including at least one NIR-band (e.g. K) with
• accuracies ≤ 0.05 mag in the optical and ≤ 0.1 mag
in the NIR
allow to disentangle ages and metallicities and deter-
mine individual GC metallicities to < ±0.2 dex,
and ages to < ±0.3 dex, i.e. they allow to distinguish
≤ 7Gyr old GCs from those ≥ 13Gyr old.
Similar studies also using NIR-data to determine
ages and metallicities of globular clusters have only
been done for a few galaxies until now (Puzia et al.
2002; Kissler-Patig et al. 2002; Hempel et al. 2003;
Larsen et al. 2005; Hempel et al. 2007).
half of these are found to host a population of GCs
that is younger and/or more metal-rich than the old
and metal-poor GC population in the Milky Way.
0.60.81.0 1.2 1.41.6 1.82.0
(F850LP - Ks)
(F475W - F850LP)
tracks of globular clusters of different metallicity in the
(F850LP − Ks) − (F475W − F850LP)
from our GALEV evolutionary synthesis models and using
AB magnitudes; Tick-marks are shown every 2 Gyrs. The
vertical lines indicate the peak-colours of the bimodal
optical colour distribution (taken from Peng et al. 2006).
We note that many of these previous studies relied on
cumulative distribution. The most important change
compared to our analysis is that we derive the physical
parameters for each individual cluster. That also en-
ables us to study correlations of these parameters, e.g.
with their spatial distributions.
We used our GALEV evolutionary synthesis models for
star clusters (Schulz et al. 2002; Anders and Fritze-v. A.
2003) to compute a large grid of models for five differ-
ent metallicities −1.7 ≤ [Fe/H] ≤ +0.4 and ages be-
tween 4Myr and 16Gyr with time-steps of 4Myr. Since
early-type galaxies do not contain significant amounts
of dust we did not include extinctions E(B − V) > 0
into our grid. We therefore only need three filters (HST
F475W,F850LP and SOFI Ks) to determine all relevant
parameters (age, metallicity, mass) for each cluster.
Note that we do not depend on color-transformation
from the HST to Standard Johnson filters. Our models
first compute spectra as function of time, that later are
convolved with the corresponding filter curves to yield
final magnitudes. Figure 1 shows the resulting color-
color diagram for the three filters discussed here.
As expected, there is a degeneracy between ages and
metallicities, but note how nicely the addition of the
near-infrared splits up different metallicities and allows
to break this degeneracy.
3.1 Near-infrared observations
We used the SOFI (Son of ISAAC) near-infrared im-
ager on the ESO-NTT during two nights in May 2007
(ESO-program id 079.B-0511). SOFI contains a Hawaii
HgCdTe 1024x1024 chip with a resolution of 0.288
arcsec per pixel, resulting in a field-of-view (FoV) of
4.92 × 4.92arcmin. This large FoV allowed us to sig-
nificantly reduce the overhead for sky-exposures using
one half of the detector for the galaxy, the other half
for sky, and changing positions every minute.
3.2 K-band data reduction
The data reduction was done using ESO-MIDAS and
largely followed the recipes given in the SOFI User’s
manual (Sterzik 2007), starting with the inter-quadrant
row-crosstalk. To flat-field the data we used a com-
bination of dome-flats, illumination correction surfaces
and a refined master-flat obtained from normalized sky-
frames of both nights. All frames were sky-subtracted
using the average of six frames taken nearest in time
that have been scaled to the sky-value of the object
exposure. We aligned individual images by matching
the positions of several stars and compact background
galaxies and averaged them to give the final image.
3.3 HST data
We used archival data obtained from the Hubble Space
Telescope. Both datasets (J8FS18011 and J8FS18021)
were observed as part of The ACS Virgo Cluster Sur-
vey (Cˆ ot´ e et al. 2004; Jord´ an et al. 2004) and automat-
ically reduced and calibrated by the On-the-fly Repro-
cessing (OTFR) pipeline at STScI. After retrieval we
checked the alignment of both frames relative to each
other, again using a set of stars within the ACS FoV.
3.4 Cluster selection and photometry
Cluster selection was done using SExtractor(Bertin and Arnouts
1996) requiring at least 4 pixels with intensities above
the 3σ-over-background threshold. A cross-correlation
of the catalogs of both HST filters to remove remain-
ing spurious detections resulted in a “optical” cata-
log of330 sources. For all these cluster candidates
we derived intrinsic radii using the ISHAPE package
within BALOAB (Larsen 1999), assuming a circular-
symmetric King profile with concentration c = 30 for
To remove background galaxies and stars, we re-
quired all valid cluster candidates to have radii in
Intermediate-age and metal-rich GCs in NGC45703
6.5 7 7.5 8 8.5 9 9.5 10
Stellar metallicity [Fe/H]
Fig. 2 Distribution of ages (upper panel) and metallicities
(lower panel) for all clusters
the range 0.2px ≤ r ≤ 5px (equivalent to physical
sizes of 0.8...20pc at the assumed distance of 17Mpc
(Tonry et al. 2001) ), leaving ≈ 280 candidates. For all
of them we performed aperture photometry with aper-
ture sizes of 10 pixels for the HST filters and 7 pixels
(≈ 2arcsec) in Ks. This allowed us to obtain (g, z, Ks)
photometry for ≈ 150 candidates; most of the remain-
ing cluster candidates for which we could not derive
magnitudes in all three filters were not included within
the SOFI FoV.
We derived physical parameters for all GC candidates
using AnalySED (Anders et al. 2004,?). It compares
the observed spectral energy distributions (SEDs) with
all model SEDs and automatically derives probabilities
for each SED on the basis of a χ2-algorithm. From
these probabilities it finds the best-fitting template and
its physical parameters age and metallicity. The results
are shown in Figure 2.
4.1 Old and intermediate ages
The age distribution is dominated by two distinct popu-
lations: An old population with ages older than 10 Gyr
that has formed during galaxy formation in the early
universe. This population is universal in all galaxies
and can be studied in great detail within our Milky
Way. The second population has intermediate ages of
1–3 Gyr and therefore must have formed later during
a violent episode of star-formation; the most plausible
explanation for such an event being an intense starburst
as e.g. accompanying the merger of two gas-rich spirals
or the accretion of a gas-rich companion, resulting in a
phase of massive star cluster formation. Besides these
two dominating populations there is a number of clus-
ters that do not belong to either of these two. These can
be explained by gas left over from the original galaxies
that was ejected into tidal tails and later rained down
onto the merger remnant.
A merger or accretion event is supported by the
detection of a nuclear stellar disk in the host galaxy
(van den Bosch et al. 1998; van den Bosch and Emsellem
1998; Scorza and van den Bosch 1998). van den Bosch and Emsellem
(1998) estimated an age of ≤ 2Gyr for the central struc-
ture, in excellent agreement with our ages for the inter-
mediate population, suggesting that both have formed
from the same event.
4.2 Solar metallicities
The lower panel of Fig. 2 shows the metallicity dis-
tribution of our cluster sample.
catching prominent peak at solar metallicity. 26 out
of 36 of these high-metallicity clusters belong to the
intermediate-age population, however, 10 out of the 36
solar metallicity GCs have old ages. Their origin is not
We caution the reader that the number of old and
metal-poor globular clusters is underestimated with our
analysis. Since we only include globular clusters with a
secure Ks-band detection, we are biased toward higher-
metallicity clusters, because those have significantly
redder optical–near-infrared colours.
A more detailed analysis of the properties of our clus-
ter sample, including their masses will be published in
Kotulla et al., in prep.
It features an eye-
We obtained deep K-band photometry of the Virgo
lenticular NGC4570 and combined it with archival opti-
cal data from HST. We selected globular cluster candi-
dates based on the HST data and their intrinsic optical
GALEV evolutionary synthesis models were used in
combination with AnalySED to automatically derive
physical parameters age, metallicity and mass for each
We detect a significant population of intermediate-
age (1–3 Gyr) and metal-rich ([Fe/H] > −0.4) cluster
population that has not been reported for this galaxy
Space Science Institute (ISSI) for their hospitality and
support of this research. This publication is based on
observations made with ESO Telescopes at the La Silla
Observatory under programme ID 079.B-0511.
paper is also based on archival observations with the
NASA/ESA Hubble Space Telescope, obtained at the
Space Telescope Science Institute, which is operated by
the Association of Universities for Research in Astron-
omy (AURA), Inc, under NASA contract NAS 5-26555.
We thank the International
Intermediate-age and metal-rich GCs in NGC45705
Anders, P., Bissantz, N., Fritze-v. A., U., de Grijs, R.: Mon.
Not. R. Astron. Soc. 347, 196 (2004)
Anders, P., de Grijs, R., Fritze-v. A., U., Bissantz, N.: Mon.
Not. R. Astron. Soc. 347, 17 (2004)
Anders, P., Fritze-v. A., U.: Astron. Astrophys. 401, 1063
Bertin, E., Arnouts, S.: Astron. Astrophys. Suppl. Ser. 117,
Brodie, J.P., Strader, J.: Annu. Rev. Astron. Astrophys.
44, 193 (2006)
Cˆ ot´ e, P., Blakeslee, J.P., Ferrarese, L., Jord´ an, A., Mei, S.,
Merritt, D., Milosavljevi´ c, M., Peng, E.W., Tonry, J.L.,
West, M.J.: Astrophys. J. Suppl. Ser. 153, 223 (2004)
Fritze-v. A., U.: Astron. Astrophys. 414, 515 (2004)
Gebhardt, K., Kissler-Patig, M.:
Hempel, M., Hilker, M., Kissler-Patig, M., Puzia, T.H.,
Minniti, D., Goudfrooij, P.: Astron. Astrophys. 405, 487
Hempel, M., Zepf, S., Kundu, A., Geisler, D., Mac-
carone, T.J.: Astrophys. J. 661, 768 (2007)
Jord´ an, A., Blakeslee, J.P., Peng, E.W., Mei, S., Cˆ ot´ e, P.,
Ferrarese, L., Tonry, J.L., Merritt, D., Milosavljevi´ c, M.,
West, M.J.: Astrophys. J. Suppl. Ser. 154, 509 (2004)
Kissler-Patig, M., Brodie, J.P., Minniti, D.: Astron. Astro-
phys. 391, 441 (2002)
Kundu, A., Whitmore, B.C.: Astron. J. 121, 2950 (2001)
Kundu, A., Whitmore, B.C.: Astron. J. 122, 1251 (2001)
Larsen, S.S.: Astron. Astrophys. Suppl. Ser. 139, 393
Larsen, S.S., Brodie, J.P., Strader, J.: Astron. Astrophys.
443, 413 (2005)
Peng, E.W., Jord´ an, A., Cˆ ot´ e, P., Blakeslee, J.P., Fer-
rarese, L., Mei, S., West, M.J., Merritt, D., Milosavl-
jevi´ c, M., Tonry, J.L.: Astrophys. J. 639, 95 (2006)
Puzia, T.H., Zepf, S.E., Kissler-Patig, M., Hilker, M., Min-
niti, D., Goudfrooij, P.: Astron. Astrophys. 391, 453
Schulz, J., Fritze-v. A., U., M¨ oller, C.S., Fricke, K.J.: As-
tron. Astrophys. 392, 1 (2002)
Scorza, C., van den Bosch, F.C.: Mon. Not. R. Astron. Soc.
300, 469 (1998)
Sterzik, M.: SOFI User’s Manual. 2.1 edn. (2007)
Tonry, J.L., Dressler, A., Blakeslee, J.P., Ajhar, E.A.,
Fletcher, A.B., Luppino, G.A., Metzger, M.R., Moore, C.B.:
Astrophys. J. 546, 681 (2001)
van den Bosch, F.C., Emsellem, E.: Mon. Not. R. Astron.
Soc. 298, 267 (1998)
van den Bosch, F.C., Jaffe, W., van der Marel, R.P.: Mon.
Not. R. Astron. Soc. 293, 343 (1998)
West, M.J., Cˆ ot´ e, P., Marzke, R.O., Jord´ an, A.: Nature
427, 31 (2004)
Astron. J. 118, 1526
This manuscript was prepared with the AAS LATEX macros v5.2.