Constraining Stellar Evolution Models using Observations of Stars in Nearby Galaxies

Rosenfield Philip, Seattle WA

Investigators

Abstract

Philip Rosenfield is awarded an NSF Astronomy and Astrophysics Postdoctoral Fellowship to carry out a program of research and education at the Harvard-Smithsonian Center for Astrophysics. The focus of his program is to improve models of stellar evolution using observations from the Hubble Space Telescope of tens of thousands of stars in other galaxies. Massive stars are largely responsible for regulating the intensity of star formation over cosmic stretches of time. Nearly any attempt to interpret observations of galaxies requires understanding the massive stars. Despite their importance, current models of massive stars do not match current observations. Historically, calibrations of massive stars models are limited to a relatively small number of stars formed in similar environments, and models of massive stars in other environments are largely extrapolations. With the recent surge in large surveys that reveal individual stars in more distant galaxies, astronomers now have the necessary data to significantly improve massive star models in a diverse set of environments. The educational program aims to increase the confidence and enthusiasm in learning astronomy of Boston Middle School students by preparing students to create and present their own digital planetarium shows. The primary goal of this program is to systematically calibrate the uncertain aspects of widely used, state-of-the-art stellar evolution models, using the best possible data as constraints across the widest range of metallicity. All new models will be publicly disseminated and Rosenfield will work closely with population synthesis modelers to ensure the models can quickly and easily be incorporated into their analysis routines. To assess which is the most probable stellar evolution model given the observational constraints, Rosenfield will calculate large sets of stellar models spanning the uncertain model parameter space as well as the metallicity range of the data. Each set will then be statistically compared to observational data. By requiring acceptable fits across all metallicities and masses, not only will the best fitting parameters become known, but any degeneracies between parameters will be revealed. For the educational program, Rosenfield will work with volunteers and teachers to prepare Boston Middle School students to give digital planetarium shows to their peers.