Markov Chain Monte-Carlo Models of Starburst Clusters
Abstract
There are a number of stochastic effects that must be considered when comparing models to observations of starburst clusters: the IMF is never fully populated; the stars can never be strictly coeval; stars rotate and their photometric properties depend on orientation; a significant fraction of massive stars are in interacting binaries; and the extinction varies from star to star. The probability distributions of each of these effects are not a priori known, but must be extracted from the observations. Markov Chain Monte-Carlo methods appear to provide the best statistical approach. Here I present an example of stochastic age effects upon the upper mass limit of the IMF of the Arches cluster as derived from near-IR photometry.
- Publication:
-
New Windows on Massive Stars
- Pub Date:
- January 2015
- DOI:
- 10.1017/S1743921314006498
- Bibcode:
- 2015IAUS..307..123M
- Keywords:
-
- Methods: statistics;
- Stars: evolution;
- atmospheres;
- rotation