Showing papers on "Mass segregation published in 1984"

Automated star counts in the southern globular-cluster ngc-6809(m55)

Abstract: THE ASTRONOMICAL JOURNAL VOLUME 89, NUMBER 1 AUTOMATED STAR COUNTS IN THE SOUTHERN GLOBULAR CLUSTER NGC 6809 (M55) MICHAEL J. IRWIN Institute of Astronomy, Madingley Road, Cambridge CB3 OHA, England VIRGINIA TRIMBLE Department of Physics, University of California, Irvine, California 92717 and Astronomy Program, University of Maryland, College Park, Maryland 20742 Received 6 June 1983; revised 23 September I 983 ABSTRACT We report star counts, as a function of position and apparent magnitude, in the rich, relatively open southern globular cluster NGC 6809 (M55). Three AAO 150q plates were scanned by the Automatic Plate Measuring System (APM) at the Institute of Astronomy, Cambridge, and 20 825 images were counted by its associated software. Previously known features of rich globular clusters which appear in the raw counts include a flattening of the luminosity function below M 3 ~ + 5.2, increased central concentration of bright stars relative to faint ones (normally interpreted as mass segregation), and mild deviations in radial profile from King models. Crowding of the field, which causes the counting proce- dure to miss faint stars preferentially near the cluster center, contributes to all of these, and may be responsible for all of the apparent mass segregation, but not for all of the other two eflqects. The deviations from a smooth radial profile are real in the sense of showing up in other counts of this and other clusters and are of marginal statistical significance (1-20 for a Poisson distribution). The luminosi- ty function is the most detailed so far published for a globular cluster, and its precise shape near main- sequence turnofi‘ may be usable as an age indicator. Automated subtraction and comparison of plate pairs rediscovered five of the six known RR Lyrae variables and found one additional candidate variable of about the same brightness, as well as eight possible variables near the faint limit of the scans. Three of these are confirmed as variable by Liller, using conventional techniques. These are of appropriate luminosity, periods, and number to be either W UMa (contact binary) or BY Dra (spotted, rotating) stars and so are deserving of further investigation. We conclude that the APM in its normal image analysis mode did at least as well in its first try at measuring stars in a crowded globular cluster field as a novice human observer would. An experienced human observer can, however, extract more reliable information in the centers of such clusters, and the APM software can probably be modified to imitate JANUARY 1984 some of the fruits of such experience. I. INTRODUCING THE APM Counting stars as a function of apparent brightness and position in the sky has been part of astronomy at least since the time of Newton (1962; Hoskin 1982). Star counts, prop- erly done, can be remarkably informative, yielding informa- tion on problems ranging from the overall structure of our galaxy (Herschel 1785; Bahcall and Soneira 1981) to the de- tailed distribution of obscuring matter (Bok 1977; Tomita et al. 1979; Trimble 1977). They can also be remarkably tedious to perform (Weistrop 1968), requiring great patience and great consistency on the part of the counter. This is particu- larly so for the crowded fields of globular clusters, where star counts remain a major source of information on cluster structure and dynamics (King et al. 1968; Harris and Racine 1979) Interesting points not yet fully clarified include the Initial Mass Function of the main-sequence stars, the extent of mass segregation within clusters, deviations of the radial profiles from classic King (1966) models, and the differences in these things from one cluster to another. Da Costa (1982) and others have begun to address these questions with star counts done by traditional methods. We here ask the addi- tional question: can we be replaced by a machine? The an- swer, perhaps predictably, proves to be a qualified yes. Her- zog and Illingworth (1977) reached a similar conclusion using a PDS scanner and a CDC 6400 computer, but their counts do not seem to have been published. The machine in our case is the Automatic Plate Measur- ing System (APM) at the Institute of Astronomy in Cam- 83 Astron. J. 89 (1), January 1984 0004-6256/84/010083-122500.90 bridge (Kibblewhite et al. 1983). It is the product of nearly a decade of SERC-sponsored development (under the leader- ship of Edward Kibblewhite) and came into regular use in 1980. Its distinctive features include a laser spot scanner, permitting very rapid scanning and low measurement noise, and on-line processing capability that includes provision for density wedge calibration, sky background estimation, noise removal, and automatic image detection and computation of image parameters (integrated isophotal intensities, posi- tions, second order moments, peak intensities, and areal pro- files). The high scanning speed of the APM enables us to take extra care in determining the sky background. All measure- ments are done in two passes. On the first pass, the sky back- ground is estimated. The plate is partitioned into roughly fix §-mm pixels, and, for each of these background pixels, an array of 64X 64 intensity measurements is made. The inter- polated mode of this array is used as an initial estimate of the sky background. When the full 2D array of initial back- ground intensity measurements has been made, it is further processed through a nonlinear filter to give the final back- ground values. The purpose of the filter is to detect and cor- rect background values contaminated by the presence of re- solved astronomical images and to smooth the final values. By making use of all the background information in this way, it is possible to estimate local background levels over extended objects such as globular clusters or nearby large galaxies. On the second pass over the plate, a threshold is defined as a fixed additive isophote above the sky background, and © 1984 Am. Astron. Soc. 83 © American Astronomical Society 0 Provided by the NASA Astrophysics Data System