Probing the Milk Way's Building Blocks: The Galactic Bulge, Stellar Halo, and Kepler Field-of-View

With the new Gaia space mission currently operational, specifically designed to map the structure of the Milky Way to unprecedented accuracies by measuring 1 billion of stars, the answers to some of the most fundamental questions of Galactic formation and evolution will be within reach in the immediate future. This thesis focuses on the preparation required for the scientific exploitation of what will be the largest and most diverse astronomical data-set ever produced, as such it is divided into two parts; the investigation of the Galactic Bulge metallicity distribution and the development, implementation, and testing of a framework to detect over-densities in multidimensional parameter space. In the first part of this thesis, additional constraints to the current galactic formation and evolution theories are provided, through the detection of Galactic Bulge RR Lyrae stars metallicity gradients in the OGLE-III sample. Using a Monte Carlo simulation to account for observational uncertainties and systematic errors, the significance and spatial variation of RR Lyrae stars metallicity gradients within the Galactic Bulge are tested on the basis of photometric data. Our results indicate that minute but statistically significant gradients exists which adds additional evidence that the Galactic Bulge may have been formed as a result of hierarchical mergers. In the second part of this thesis, we prepare for the exploitation of what will be the gold-mine data-set for galactic formation and evolution by implementing and testing a hierarchical density based clustering algorithm named OPTICS, and its fuzzy counter-part FOPTICS, to detect overdensities in multi-dimensional parameter space. We have analyzed the performance of OPTICS/FOPTICS on two test applications, which mimic use possibilities for the Gaia mission data exploitation and future space missions such as TESS and PLATO. To show the power and versatility of OPTICS/FOPTICS, we have tested its performance as a stellar halo structure finder using a synthetic stellar halo from Bullock and Johnston (2005) as well as an unsupervised stellar variability classifier using the NASA Kepler satellite stellar light curve data. In the process OPTICS has been optimized to achieve a near linear run-time complexity, which makes the method suitable for very large datasets.status: publishe