Galactic Archaeology: the Milky Way in the context of large scale surveys

Abstract

In this thesis, I develop a flexible isochrone fitting framework to derive stellar parameters using either spectroscopic or photometric input data, with the goal of applying this machinery for studies of Galactic Archaeology. Depending on the input (spectroscopic/photometric/astrometric), my machinery outputs inferred stellar parameters (e.g., effective temperature, surface gravity, metallicity, age and mass) together with their uncertainties given various of Bayesian priors. I apply this framework to observations from the GALAH survey to derive stellar ages and investigate the temporal evolution of elements and its implications into the formation of the thin and thick disks. I then adapt this framework into deriving photometric stellar parameters, distances and extinctions. This allows me to derive reliable metallicities for 10**7 stars in the SkyMapper Southern Survey, increasing the usability of this survey for studies of Galactic Archaeology.

In the first chapter of this thesis, I detail the construction of the Bayesian framework of isochrone fitting in deriving stellar ages, incorporating absolute K magnitude (M_K) derived from 2MASS photometry, Gaia data release 1 (DR1) parallax and measured stellar parameters. As a test, I recompute stellar ages and masses of ~4000 stars in the solar neighborhood from six well-studied literature samples using both Hipparcos and Gaia DR1 parallaxes. The ages are found to be compatible with literature values but with reduced uncertainties in general. I also reconstruct the local age-metallicity relationship (AMR) by recomputing ages for approximately 3000 stars in the Geneva-Copenhagen Survey. I find the reconstructed AMR to contain smaller scatters for all ages up to 12 Gyr compared to literature values.

In the second chapter, I refine and deploy this algorithm on ~160,000 turn-off and sub-giant stars from the GALAH Survey. The second data release of the GALAH Survey contains the largest existing high-resolution spectroscopic sample, which provides de- tailed elemental abundances for several hundred thousand stars up to 3 kpc. I find the local AMR to be nearly flat but with significant scatter at all ages. After correcting for selection effects, the AMR appears to have intrinsic structures indicative of two star formation events, which could be connected to the thin and thick disks. I also present abundance ratio trends for 16 elements as a function of age, across different metallicity bins. In particular, the r-process element Eu shows decreasing [Eu/Fe] with time, favouring primary production from sources with a short time-delay such as core-collapse supernovae over long-delay events such as neutron star mergers.

In contrast to the high-resolution, spectroscopic surveys like GALAH, photometric surveys on average offer orders of magnitude more stars. To exploit this larger sample, I tailor my algorithm to the SkyMapper survey, deriving photometric distances, extinctions and stellar parameters for 17 million stars from SkyMapper DR3. This work is presented in the third chapter. I use a total of 12 photometric bands across three surveys (SkyMapper, Gaia and 2MASS) in addition to priors on parallaxes and interstellar extinction. I show the quality of my stellar parameters with a case study of the Gaia red and blue sequence. I find their metallicities agree with the interpretation that the blue sequence being composed by (accreted) halo stars, whereas the red sequence is the kinematically hot tail of the thick disk, thus demonstrating the potential SkyMapper bears for stellar population studies. This sample will serve as a foundation for my future work into the use of kinematic information for studies of Galactic Archaeology. Show more