next up previous contents index
Next: Combined IRAC2 vs Up: StarsStellar Populations, Previous: The M87 Globular

Science with the Hubble Space Telescope -- II
Book Editors: P. Benvenuti, F. D. Macchetto, and E. J. Schreier
Electronic Editor: H. Payne

M3: An Ideal Laboratory for Testing Stellar Evolution and Dynamics

F. Fusi Pecci, R. Buonanno, D. Burgarella, C. Cacciari, C.E. Corsi, B. Dorman, G.G. Fahlman, F.R. Ferraro, M. Laget, G. De Marchi, G. Marconi, B. Paltrinieri, F. Paresce, H.B. Richer, R.T. Rood

Osservatorio Astronomico, Bologna, Italy
Osservatorio Astronomico, Roma, Italy
Lab. Astronomie Spatiale, CNRS, Marseille, France
Goddard Space Flight Center, Greenbelt, MD, USA
University of British Columbia, Vancouver, Canada
European Southern Observatory, Germany
Università di Bologna, Italy
University of Virginia, Charlottesville, VA, USA



Preliminary reductions of WFPC2 frames have yielded new Blue Straggler candidates in the very central region and a candidate binary sequence to be checked further.

Keywords: globular clusters,M3,photometry,HST


The basic aims of this program are:

  1. Perform the most accurate test of the ``stellar clock'' used in dating globular clusters.
  2. Obtain a complete survey of very fast evolutionary stages and rare stellar types (e.g., UV-bright stars, blue stragglers, binaries) along with their radial distributions; this will provide information on the internal cluster dynamics and on the origin of UV light in elliptical galaxies.

M3 is the ideal cluster for this purpose: we already have ground-based photometry of about 18,000 stars at radial distance providing the best complete CMD (over ), the largest complete Luminosity Function for post-main sequence stars, and the largest sample of blue stragglers ever obtained for a globular cluster from the ground (Buonanno et al. 1994).

The tests we want to perform require even larger samples, which can only be obtained by observing the cluster core, not resolved from the ground, hence the need for HST. In addition, it's important to derive the radial dependence of various properties and stellar types right to the center of the cluster.

These data will provide the first complete sample of post-main sequence stars over the entire area of the cluster, and will allow the most stringent tests of stellar evolution and dynamics models.

Observations and Data Reduction

The Observations

WFPC2 observations (CYC4- GO5496) of the galactic globular cluster M3 have been taken in U (F336W), V (F555W), I (F814W), and the mid-UV filter (F255W), mapping three positions: the cluster core and two external positions at radial distance 78 and 156 arcsec.

The Color-Magnitude Diagrams presented here are preliminary results on two fields only, i.e., the PC located on the cluster center, and the adjacent WF3 field.

The exposures taken in each filter are: F255W: 4 300sec; F336W: 4 800 sec + 2 70sec; F555W: 4 100 sec + 2 3sec; F814W: 4 140 sec 4 + 2 3sec.


In order to make a master unsharp mask image for each filter, all the long-exposure images were registered using MIDAS, yielding a median frame while statistically eliminating cosmic rays.

The search of the individual star components was made on the median frame following the usual procedure using two different packages (ROMAFOT and DAOPHOT). Then, we performed the PSF fitting on each individual frame separately and computed the average instrumental magnitudes for each star.

Preliminary Results

Candidate Binaries

An intriguing feature, clearly visible in the CMD in Figure 1a, is a sequence of objects running parallel to the Main Sequence (MS) on its right-hand side. Photometric errors and, possibly, contamination by residual unresolved objects make it difficult to separate unambiguously the parallel sequence from the MS.

As well known, binary systems (both ``physical'' and ``optical'') formed by two identical objects are expected to populate a sequence parallel to the MS and mag brighter. Binaries formed by two components having different luminosities (masses) are expected to fill in the region between the MS and the brighter envelope.

In order to find better evidence of the existence of the parallel sequence and derive a preliminary estimate of the binary frequency in this annular region located at 50" < r < 100" from the adopted cluster center, we have selected two magnitude intervals where we perform a simple test. We have defined a MS ridge-line by eye, and refined it by an iterative procedure that discarded all stars deviating by more than 3. Then we have obtained the color distributions of all the stars in the two selected magnitude ranges with respect to this MS ridge line. The observed distributions are quite clearly bimodal. By using a simple fit based on the convolution of two distinct Gaussians, one gets a preliminary estimate of the average binary fraction of about 10--15.

We, therefore, conclude that a significant number of binary candidates is present in the observed field and in the quoted magnitude interval.

Blue Stragglers

M3 contains the largest population of Blue Straggler Stars (BSS) detected in a globular cluster so far (). Moreover, from the analysis of a complete sample of bright BSS (B) Ferraro et al. (1993) showed that the radial distribution of the BSS in M3 is not monotonic, but shows a dip between 4() and 8().

Based on the new PC data, we have detected 71 BSS candidates (filled dots in Fig. 1b) in the very central region of the cluster, most of them fainter than B=18.6. Identifications with any previous lists are in progress, but it is already evident that the total number of BSS candidates in this area has been significantly increased by the present study.

To estimate the impact of these new detections on the relative frequency of BSS in M3 as a function of radial distance, we have selected the samples of BSS and comparison SGB stars in the same area and magnitude range, following the same procedure used by Ferraro et al. (1993). Then, we have computed the doubly normalized ratios and , where N is the star number and L is the luminosity.

The relative frequency of the BSS so obtained for the most internal region from the present PC data is homogeneous and compatible with the more external data-points previously derived from ground-based observations. It shows up quite clearly that the central area contains a larger normalized number of candidates; this adds further support to the evidence that the central region of M3 is an ``ideal'' environment for the BSS.

Setting aside any discussion about the various mechanisms able to produce the BSS, the possible detection of binaries in the central regions of this cluster would strengthen the connection between the BSS and the population of binary stars.

Figure: Fig. 1: HST Instrumental CMDs of M3


We gratefully acknowledge financial support from the Agenzia Spaziale Italiana (ASI).


Buonanno, R., Corsi, C.E., Buzzoni, A., Cacciari, C., Ferraro, F.R., & Fusi Pecci, F. 1994, A&A, 290, 69

Ferraro, F.R., Fusi Pecci, F., Cacciari, C., Corsi, C.E., Buonanno, R., Fahlman, G.G., & Richer, H.B., 1993, AJ, 106, 2324

next up previous contents index
Next: Combined IRAC2 vs Up: StarsStellar Populations, Previous: The M87 Globular