The ACS Globular Cluster Survey

The Project

Figure 1: Globular Cluster 47 Tucanae; Credit: W. Keel (U. Alabama in Tuscaloosa) et al., 1.54-m Telescope, Chile

The Globular Cluster Treasury program (PI: Ata Sarajedini, University of Florida) is an imaging survey of Galactic globular clusters using the ACS/WFPC instrument on board the Hubble Space Telescope.

Why Globular Cluster?

Globular clusters (GCs) are powerful tracers of the Galactic halo, thick disk, and bulge populations. The ages, metallicities, and kinematics of the GCs bear the imprint of the Galaxy's early formation history. Cluster stars- sharing the same age and chemical composition and only differing in mass- play an important role in studies of stellar structure and evolution and in the formation and evolution of the Galaxy. Cluster luminosity function scan yield the Initial Mass Function (IMF) if stellar evolution and dynamical effects are taken into account. Combined, the Galactic GCs have the potential to answer questions in a broad range of fields from star formation and evolution to galaxy formation and evolution. However, this requires an extensive AND uniform database of photometry and astrometry for individual stars in clusters with a range of properties.

Figure 2: Globular Cluster M22 from CFHT; Credit & Copyright: Jean-Charles Cuillandre (CFHT), Hawaiian Starlight, CFHT

HST already contributed tremendously to this study, focusing on specific clusters and conducting general surveys. In the largest systematic survey – the "HST snapshot survey"- data for 74 Galactic GCs were obtained in the F439W and F555W filter bands. A catalog of CMDs for the cluster sample is available from Piotto et al. (2002) .

Based on the performance of the WFPC2 instrument the HST/WFPC2 snapshot survey provides reliable photometry only for stars brighter than the turnoff. ACS extends the WFPC2 data set in three fundamental ways:

  1. The observations allow a systematic study of both lower main sequence stars, to 0.2 Msolar with S/N ≥ 10, and the brightest degenerates ( ≥ 0.5 mag) on the white dwarf (WD) cooling sequence. The ACS dataset will be used to:

          1. determine distances via comparison with local sub-dwarfs;

          2. measure relative and absolute GC ages;

          3. obtain the mass function of the central regions, and provide an observational measurement of the mass segregation in the inner regions of GCs

  1. Taking advantage of previous WFPC2 images, we will measure proper motions of the brightest stars in key clusters. This allows us to:

          1. separate cluster members from field stars, particularly for contaminated disk/bulge clusters

          2. determine proper motion dispersions in order to study their internal dynamics

          3. determine distances by comparing the proper motion and radial velocity distribution

          4. determine absolute proper motions for GCs with respect to background galaxies and hence 3-D velocities and orbits

          5. probe the structure and kinematics of the Milky Way

          6. provide synergy with observations from the Space Interferometry Mission (SIM)

The Method

Cluster Ages, Distances and Metallicities

Globular clusters are the oldest Galactic objects for which reliable ages can be measured. Their absolute ages place an upper limit on age of the Universe. Their relative ages place fundamental constraints on the mechanisms by which the Galaxy formed. The debate over the relative importance of the monolithic collapse, corresponding to a single age stellar population (Eggen, Lynden Bell, and Sandage 1962) , and accretion, associated with a broad range of ages (Searle & Zinn 1978) as Galaxy formation scenarios is still ongoing and globular clusters play an important role in this debate.

Using globular clusters to set constraints on the formation and evolution of the Galaxy requires absolute ages for a sample of clusters covering the entire GC metallicity range, as well as relative ages for as many globular clusters as possible, distributed all over the Galaxy, and again, covering the entire metallicity range of Galactic GCs.

Figure 3: 47 TUC CMD (Payne-Gaposchkin & Haramundanis 1970)

To determine absolute and relative ages as well as metallicities for Galactic GCs requires the location of the Main Sequence Turnoff (MSTO) with high accuracy for many GCs. Photometric homogeneity is a fundamental requirement. The ACS data in our survey provide extremely well defined CMDs for a large sample of GCs extending from the tip of the Red Giant Branch (RGB) to at least ~6.5 mags below the MSTO, with the highest photometric accuracy around the TO. In the CMD we can determine the position of the TO with unprecedented precision. Based on the dependence of cluster ages on metallicity, cluster morphology, HB morphology, kinematics and position within the Galaxy we can now study the early formation history of the Milky Way.

To determine precise relative ages for clusters of similar metallicities the classical techniques of the difference in color between the MSTO and the RGB, and the difference in magnitude between the Horizontal Branch (HB) and the MSTO (Sarajedini & Demarque 1990, VandenBerg et al. 1990). The absolute magnitude of the TO and the sub-giant branch define the absolute ages of the clusters which will allow age comparisons between clusters of different metallicities (see Figure 3).

However, for accurate absolute age estimates accurate distances, reddenings, and metallicities are essential. The main source of error in absolute age measurements is the distance. Depending on the actual target cluster we adopt one of two different methods of distance estimation. For some the distances, with an uncertainty of only a few percent, will be derived from proper motion dispersion, based on multi-epoch HST images, and radial velocity dispersion. For a number of GCs we will use WFPC2 images, available from the HST snapshot survey, to determine the proper motion dispersion. In addition we will also use high resolution spectroscopy (obtained with FLAMES @ VLT (ESO)) , MS fitting to field sub-dwarfs with Hipparcos parallexes.


As a starting point for reddening corrections we rely on the reddening maps by Bustein & Heiles (1982), Burstein (2003) and Schlegel (1998) . These will be augmented by the results of simultaneous reddening and metallicity technique developed by Sarajedini and Saviane. Ongoing ground based spectroscopic surveys (e.g. by Carretta et al. 2003) will provide additional reddening and metallicity values.

Eventually we will combine these preexisting data with the high accuracy CMDs obtained in this program and derive the distance, age, reddening, and metallicity for our globular cluster sample. The final uncertainty of the absolute ages, mostly caused by model uncertainties, will be in the range of 1 Gyr.

The Stellar Mass Function

If we are to understand the early history of the Milky Way and/or galaxy formation in general it is necessary to know the stellar mass function at low metallicities, for which GCs are an ideal probe. Previous studies have been hampered by three main issues: image crowding (i.e. contamination with field stars) and radial mass segregation. The GC Treasury program addresses those problems by answering the following questions:

Proper Motions

One of the major goals of GC Treasury program is to obtain proper motions of stars in the cluster fields. The later are essential to clean the CMDs from contaminating field stars.

As part of the program we will derive the proper motions for nearby clusters with sufficient accuracy to yield the internal motions of clusters and allow contemplation of a number of studies involving internal cluster dynamics, including:

The database built in this program is a valuable asset to the Space Interferometry Mission ( SIM). Although the results of the GC Treasury program can not compete with the SIM project with respect to precision (micro-arcseconds) the later is hampered by the small sample size, typically three stars per cluster. Additionally, only the 32 closest GCs are included in SIM. The internal motions derived from the extensive ACS data permit estimation of the relative motion of the SIM-measure stars with respect to the bulk cluster motion. Our survey also improves the efficiency and accuracy of SIM by separating cluster target stars clear of nearby (<1.0”) companions that would modulate the primary interferometry fringes.