Massive stars are often found embedded in a galaxy’s dusty interstellar medium. If there is any dust present along a line-of-sight, the light from stars will be reddened, creating a degeneracy, in which the spectra energy distribution (SED) of a heavily-extinguished massive star might end up looking similar to a cool star with low dust. But, by fitting both a stellar model and dust extinction, we can break this degeneracy and correctly classify heavily-extinguished massive stars.
We investigate the most massive stars in the Andromeda Galaxy (M31) by characterizing Panchromatic Hubble Andromeda Treasury (PHAT) data with the Bayesian Extinction and Stellar Tool (BEAST), an open-source, open-development tool designed to simultaneously fit stellar SEDs and dust extinction. PHAT is an enormous multi-band survey that used the Hubble Space Telescope to observe roughly a third of M31, producing the photometry measurements for more than 100 million stars. With the help of the BEAST, we obtain initial mass estimates for over 45 million PHAT sources with detections in at least 4 bands. By positionally cross-match our PHAT results with high-mass spectroscopic results from Massey et al. 2016, we were able to identify several extinguished candidate O and B-stars with initial mass estimates larger than 50 M☉.