The Extremely Young Star Cluster Population In Haro 11
aa r X i v : . [ a s t r o - ph . C O ] A ug **FULL TITLE**ASP Conference Series, Vol. **VOLUME**, c (cid:13) **YEAR OF PUBLICATION****NAMES OF EDITORS** The Extremely Young Star Cluster Population In Haro 11
Angela Adamo , G¨oran ¨Ostlin , Erik Zackrisson , & Matthew Hayes Department of Astronomy, Stockholm University, AlbaNova, SE-10691 Stockholm, Sweden; Observatoire Astronomique de l’Universit deGenve, 51, ch. des Maillettes,CH-1290 Sauverny, Suisse
Abstract.
We have performed a deep multi-band photometric analysis of thestar cluster population of Haro 11. This starburst galaxy (log L
F UV = 10 . ⊙ )is considered a nearby analogue of Lyman break galaxies (LBGs) at high redshift.The study of the numerous star clusters in the systems is an effective way toinvestigate the formation and evolution of the starburst phase. In fact, theSED fitting models have revealed a surprisingly young star cluster population,with ages between 0.5 and 40 Myr, and estimated masses between 10 and 10 solar masses. An independent age estimation has been done with the EW(H α )analysis of each cluster. This last analysis has confirmed the young ages of theclusters. We noticed that the clusters with ages between 1 and 10 Myr show aflux excess in H (NIC3/F160W) and/or I (WFPC2/F814W) bands with respectto the evolutionary models. Once more Haro 11 represents a challenge to ourunderstanding. The interactions and/or mergers between galaxies are processes that usually en-hance the star formation rate (SFR) (see other contributions in these proceed-ings) and produce many luminous star clusters (SCs). The cluster formationand evolution is strictly connected with the evolutionary history of the hostgalaxy. We have focused our studies on the local blue compact galaxy (BCG)Haro 11 (ESO 350-IG 038), with estimated distance of around 82 Mpc. It showsa complex morphology, with three active and very luminous starburst regionsand irregular kinematics ( ¨Ostlin et al. 2001, ¨Ostlin et al. in prep.). A photomet-ric study of the halo of Haro 11 (Micheva et al. 2009) confirms that the stellarpopulation in the outskirts of the system is consistent with a ≥
10 Gyr old popu-lation and low metallicity. The coexistence of old stellar populations, active starforming regions (Hayes et al. 2007), and unrelaxed kinematics indicates that theHaro 11 starburst could be the result of a merger between an evolved systemand a gas-rich dwarf galaxy. As a consequence of this merging a new populationof SCs has been formed. In Figure 1 we show the image of the galaxy in theWFPC2 filter, F606W. The starburst region is dominated by the numerous SCs.Our aim is to investigate the SC properties to constraint the starburst age andpropagation. 1
Adamo A.
Figure 1. WFPC2/F606W image of Haro11. The galaxy shows three star-burst regions with numerous clusters. The central knot, where a white circleis located appears obscured by some dust lines. Inside the circle there is thebrightest cluster of Haro 11. The line shows the position of the brightestyoung cluster.
A multi-wavelength photometric analysis (from far-UV to IR), have been per-formed on the SCs. About 300 confirmed cluster candidates have been detectedin the high-resolution HST images (see Adamo et al. 2009). We show here theresults of the SED (spectral energy distribution) analysis (Adamo et al., inprep.). The used models are based on the Zackrisson et al. (2001) populationsynthesis code, which includes nebular emission from the photoionized inter-stellar medium (important when analyzing observations of young stellar popu-lations; e.g. Anders & Fritze-v. Alvensleben 2003; Zackrisson et al. 2008). Aninstantaneous burst, a Salpeter IMF and a metallicity of Z= 0 .
004 has been usedas input. In Figure 2 (left panel), we show an example of the fit. When we try tofit the H and I bands we obtain a poor fit of the blue side of the spectrum (moresensitive to age). In this case we obtained an age of 35 Myr. If we only fit thewavelength range from UV (ACS/F140LP) to V (WFPC2/F606W), excluding Hand I, we get a better fit of the blue side. The age estimated in this second caseis of 3.5 Myr. Comparing the two different spectral models with the observedfluxes (Figure 2) we see a clear excess in H and I bands that quite considerablyinfluences the age estimation (and masses) of the SCs. In order to get an in-dependent estimation of the ages of the SC we used also EW(H α ) a sensitiveestimator at young ages. However, we have to consider the ages estimated fromH α as an upper limit to the real ages of the clusters. In fact, we don’t know Cs in Haro11 α emitters. Figure 2 (right panel) shows a comparisonbetween the ages estimated from SED fit (H and I included) and EW(H α ). Theages estimated from EW(H α ) are between 1 - 35 Myr, much lower than the SEDfit outputs (1 - 100 Myr). Because of the unclear origin of this red excess atthe youngest ages Reines et al. (2008) we cant produce any ad-hoc model to fitthese young clusters. Finally we decided to exclude the H and I from the chisquare fits. In Figure 3 we can see the present mass as function of the ages of Figure 2. Left plot: the best SED fit to the detected fluxes (filled dots).The squares connected by a dotted line show the best fit model to all thefilters. The triangles connected by a dashed line is the best fit to the blue sideof the spectrum, excluding the I and H bands from the fit. The red excessis seen as an excess in flux at the redder wavelength with respect to the bestfit model. Right plot: the cluster age estimated from EW(H α ) versus theages obtained from the SED fit of all the detections (I and H included). Thedashed line indicates the region of the plain where the ages estimated withthe two method agree. The dotted lines outline the region of the plain wherethe cluster ages is between 1 and 10 Myr. the SCs in Haro11. The clusters appears young, with ages between 1 and 40Myr and masses between 10 and 10 M ⊙ . We tried to estimate the evolution of the star formation rate (SFR) in Haro 11,using the relation between the SFR and the luminosity of the brightest youngcluster in the galaxy Bastian (2008). From the estimated Haro 11 far infraredluminosity, L
F IR
Hayes et al. (2007), we know that the present SFR is around 18M ⊙ yr − . Applying the relation by (Bastian 2008), we obtain from the brightestyoung cluster (M V = − . ⊙ yr − , in perfect agreement with FIR estimations.However, the brightest cluster in the galaxy has M V = − . ⊙ yr − . This Adamo A.
Figure 3. The present mass as function of age of the clusters. The linesare the predicted detection limits using our stellar evolutionary models, forM V =28.0 mag (dotted) and M I =28.0 mag (dashed). suggests that after a first effective burst the star formation has propagated inthe galaxy with less efficiency possibly caused by feedbacks and outflows. The analysis of the SCs in Haro 11 has confirmed the young age of the ongoingstarburst in the system. The starburst involves clusters with mass range ofseveral orders. The SED analysis shows that a subsample of clusters have asignificant flux excess in H and I with respect to the predicted stellar populationmodels. The origin of this red excess is not still clear and is impossible to matchwith the usual stellar evolutionary models used so far. We will discuss our resultsin forthcoming papers.
Acknowledgments.
A. Adamo thanks the organizers for the interestingconference. We acknowledge support from the Swedish Research Council (Veten-sapsr˚adet).