The members of the star formation group Shuyu and Nils joined the PHANGS collaboration. Congratulations!
Shuyu recently developed an automatised bubble detection algorithm for supernova-driven bubbles using simulation data of isolated Milky Way-like galaxy analogues (in prep.). Shuyu is interested in supernova-driven bubbles and will be involved in linking simulation results with JWST observations for bubble studies, with a view to improving identification methods and result interpretation.
Nils works on inferring properties of star clusters in the PHANGS catalogue. This is notoriously tricky due to our position inside of the galaxy, along with dust and crowding making it hard to get a complete picture. Using self-supervised models that extract data points a mathematical “representation spaces”, it bypasses the need for potentially biased labels, allowing us to explore the data based purely on its observed properties. This robust foundation can then be fine-tuned for specific tasks (like age determination) using a small number of high-fidelity labels, often outperforming fully supervised models trained from scratch. Once fine-tuned, adaptation to numerous different analyses across large projects such as PHANGS these foundation models become a shared, powerful tool for the entire community.
The PHANGS collaboration (Physics at High Angular resolution in Nearby GalaxieS) is a research program targeting high resolution observations of nearby galaxies with several telescopes. It aims to understand the interplay of the small-scale physics of gas and star formation with galactic structure and galaxy evolution. Observations of nearby galaxies will be utilised to understand how physics at or near the “cloud” scale are affected by galaxy-scale conditions, how they affect still smaller scale processes, and how these influence the evolution of whole galaxies.