My recent work has focused on evolved low-mass stars. Are our stellar models accurate? Do they agree with new data on stellar masses, metallicities, temperatures, and rotation rates? What can the disagreements tell us about the internal physics of stars? These questions are important because any problems in stellar models become magnified when used to determine the properties of planets or infer how galaxies evolved.
I'm currently using data collected from large surveys like Kepler and APOGEE and making stellar models using the YREC stellar evolution code. My thesis is focused on using the surface and core rotation rates of red giants to probe the physics of angular momentum transport and loss, but I'm also involved in projects on stellar mixing lengths and [C/N] ratios. Additionally, I've done previous projects on young stars, using both spectra and photometric variability to probe such processes as accretion, star spots, and interactions with disks.
Stellar evolution models are used in a wide variety of fields, from planetary formation to galactic evolution. However, most of these models are calibrated assuming that every star is like the sun, and in the past astronomers lacked sufficient data to check that assumption. With the combination of asteroseismology and high resolution spectroscopy, there are now thousands of stars with known masses, radii, surface gravities and temperatures across a range of metallicities. When I compare their properties to model predictions, I find that the models can be off by hundreds of Kelvin, in a metallicity dependent way, which can be corrected if the convective mixing length is allowed to be a function of metallicity (see figures below). As new precise measurements are made in all regimes, it is imperative that we continue to check and improve our models.
Temperature offset from the YREC grid of models as a function of metallicity for the APOGEE-Kepler first ascent red giant branch sample. Note that the PARSEC grid of models has a similar slope, but a different normalization. These offsets can be corrected if the convective mixing length is set to be metallicity dependent, although the required values contradict predictions of three-dimensional simulations. See Tayar et al. (2017) for more details
Rotation is one of the biggest unknowns of stellar evolution theory. In the past, for lack of data, people assumed that all stars rotated as solid bodies all the time, and underwent angular momentum loss from magnetized winds with some calibration based on a few clusters. With the collection of core and surface rotation data for old main sequence stars and red giants, it has become clear that these simple parameterizations are insufficient. Using evolved stars, I've found that in general cores rotate faster than surfaces, that the coupling between the two depends on mass and evolutionary state, that there is more angular momentum loss that standard Kawaler-type models predict, and that there might even be radial differential rotation in the convective zones of these stars.
Left: Comparison of the predicted surface rotation rates for stars in the secondary clump assuming a Kawaler wind loss law and solid body rotation(bottom) to the measured rotation rates from Tayar et al. (2015) and Ceillier, Tayar et al. (2017) (top). Right: Comparison of various limiting case core-envelope coupling scenarios to the measured core rotation rate of KIC 7341231 (red diamond, Deheuvels et al. 2012), suggesting weak post-main-sequence coupling (Tayar & Pinsonneault 2013).