About the project
In modern Galactic astronomy, stellar spectroscopy has a pivotal role in complementing large photometric and astrometric surveys, such as Gaia, PLATO and TESS. Spectroscopic observations provide crucial data on stellar parameters, chemical compositions, and radial velocities, enabling deeper insights into the chemical evolution and chemo-dynamical mechanisms of the Milky Way and its satellites. Several large spectroscopic surveys have already provided data for millions of stars, with many more underway, promising to significantly expand our understanding of the formation and the evolution of our Galactic environment.
Despite the wealth of data from these surveys, systematic differences in derived spectroscopic parameters raise challenges. Efforts to harmonize these surveys onto a common scale are essential to maximize their scientific legacy. Machine learning techniques offer promising avenues for homogenizing spectroscopic surveys on the same base, but they require addressing issues such as parameter space coverage of the training set and the compatibility of different survey methodologies. Additionally, the creation of benchmark catalogues and the development of common metrics are critical steps in evaluating and improving homogenization methods.
The project brings together experts with different backgrounds to tackle these challenges collaboratively. Through discussions and collaborative efforts, the team aims to establish a comprehensive understanding of the homogenization process and develop new methodologies to ensure the compatibility and accuracy of spectroscopic surveys. By defining the applicability domain of homogenization methods and developing common metrics, the project aims to provide valuable guidance for future spectroscopic surveys, maximizing their scientific impact and ensuring their seamless integration with other surveys.