POLAR: Computational Intelligence for Multimodal Astrophysical Tomography
Dates: 07/2021 – 12/2024
Funded by: 3rd Call for FORTH Synergy Grants proposals
Project Coordinator: Georgios Tzagkarakis (FORTH-ICS)
PI for FORTH: Georgios Tzagkarakis
Funding: total cost/funding: € 73,600, FORTH-ICS funding: € 50,300.
Summary:
A detailed mapping of the Galactic magnetic field would revolutionize our understanding for a wide range of astrophysical processes, including star formation, high-energy astrophysics, and cosmology. Typically, our diagnostics of astrophysical magnetic fields are few, hard to obtain, and difficult to process. The main method exploits starlight polarization, induced by the magnetic field that permeates interstellar dust clouds between the stars and the Earth, as a major diagnostic of the Galactic magnetic field properties. The polarization of stars that are distributed at different, but known (thanks to ESA’s Gaia mission) distances along the line of sight, encode information on the 3D geometry of the magnetic field of the Milky Way. The problem is knowing the distance to the stars and their polarization to create a tomographic map of the magnetic field of the Galaxy. A second major problem in such studies is that the observed starlight polarization does not always encode the magnetic field of the Galaxy, but instead, an appreciable fraction of stars emits polarized light themselves. These intrinsic polarizations are a nuisance for Galactic magnetic field studies, but also a real treasure for stellar astrophysics. POLAR aspires to pave the way for the upcoming Big Astrophysical Data era by providing a solid computational intelligence framework for unsupervised (or semi-supervised) multimodal data analysis and learning in astrophysical tomography.
Visit Website