Determining open cluster membership
A Bayesian framework for quantitative member classification
Fairfield University, Department of Physics, Fairfield, CT 06824, USA
Received: 21 March 2016
Accepted: 12 September 2016
Aims. My goal is to develop a quantitative algorithm for assessing open cluster membership probabilities. The algorithm is designed to work with single-epoch observations. In its simplest form, only one set of program images and one set of reference images are required.
Methods. The algorithm is based on a two-stage joint astrometric and photometric assessment of cluster membership probabilities. The probabilities were computed within a Bayesian framework using any available prior information. Where possible, the algorithm emphasizes simplicity over mathematical sophistication.
Results. The algorithm was implemented and tested against three observational fields using published survey data. M 67 and NGC 654 were selected as cluster examples while a third, cluster-free, field was used for the final test data set. The algorithm shows good quantitative agreement with the existing surveys and has a false-positive rate significantly lower than the astrometric or photometric methods used individually.
Key words: open clusters and associations: general / methods: data analysis / methods: statistical
© ESO, 2017