Kuldeep Verma, Robert J. J. Grand, Víctor Silva Aguirre, Amalie Stokholm
An observational testbed for cosmological zoom-in simulations: constraining stellar migration in the solar cylinder using asteroseismology
See arXiv version
19 pages, 18 figures (including 8 in the appendix), 1 table, MNRAS in press
Abstract
Large-scale stellar surveys coupled with recent developments in magneto-hydrodynamical simulations of the formation of Milky Way-mass galaxies provide an unparalleled opportunity to unveil the physical processes driving the evolution of the Galaxy. We developed a framework to compare a variety of parameters with their corresponding predictions from simulations in an unbiased manner, taking into account the selection function of a stellar survey. We applied this framework to a sample of over 7000 stars with asteroseismic, spectroscopic, and astrometric data available, together with 6 simulations from the Auriga project. We found that some simulations are able to produce abundance dichotomies in the \([{\rm Fe}/{\rm H}]-[\alpha/{\rm Fe}]\) plane which look qualitatively similar to observations. The peak of their velocity distributions match the observed data reasonably well, however they predict hotter kinematics in terms of the tails of the distributions and the vertical velocity dispersion. Assuming our simulation sample is representative of Milky Way-like galaxies, we put upper limits of 2.21 and 3.70 kpc on radial migration for young (\(< 4\) Gyr) and old (\(\in [4, 8]\) Gyr) stellar populations in the solar cylinder. Comparison between the observed and simulated metallicity dispersion as a function of age further constrains migration to about 1.97 and 2.91 kpc for the young and old populations. These results demonstrate the power of our technique to compare numerical simulations with high-dimensional datasets, and paves the way for using the wider field TESS asteroseismic data together with the future generations of simulations to constrain the subgrid models for turbulence, star formation and feedback processes.
