Big Data with the Galah Survey
Galactic Archaeology with Hermes (Galah) is a high-resolution optical stellar survey that will observe up to 1 million spectra. The primary science goals of Galah are to carry out the chemical tagging experiment (finding stars of common birthplaces via their chemical composition) and to use the observational data to test theoretical stellar and galactic evolutionary models.
Our group is participating in the analysis of the large spectroscopic data set by estimating parameters and up to 30 different elemental abundances for each observed star. Because of the large amount of data, our group is using a new approach, developed at MPIA, to estimate those parameters as accurate and precise as possible, while minimising the cost of classical (time-consuming) spectrum analysis. The analysis is based on a combination of the classical method of spectrum synthesis (for a representative set of spectra with external information e.g. from seismology and/or astrometry/photometry) and the data-driven approach of The Cannon.
Because of the large information content of the survey, our group is working on the optimisation of the spectrum analysis (adding latest information on stellar atmosphere models and non-LTE behaviour of elements) as well as the exploitation of this high-dimensional chemical phase space in combination with the dynamical information by the astrometric mission of the Gaia satellite.
For more details contact Sven Buder and Karin Lind