Wednesday

Most galaxies in the local Universe divide into two groups: star-forming spirals and quiescent early-types. However, the relationship between a galaxy's star formation properties and structure is still not well-understood, and studying it requires detailed analysis of faint and small-scale galaxy features. In particular, analysis of faint post-merger features, such as tidal tails, shells, and internal disturbances, is essential to determine whether the morphological evolution is merger-driven. Deep and wide surveys, such as Vera Rubin Observatory, will enable studying these important features in large galaxy populations.

However, wide surveys have limited resolution, while the data volume makes morphological studies computationally challenging. I will present our recent findings that high-resolution imaging is essential to detect small-scale disturbances, and that over 80% of transitioning galaxies are morphologically disturbed in HST imaging but not in SDSS. To remedy this, we devised a machine learning model that 1) speeds up morphological analysis and 2) corrects for noise and resolution effects. Our new model aims to provide consistent morphology measurements with high-resolution HST imaging and lower-resolution wide surveys, and will enable robust morphological analyses of unprecedented data volumes provided by the Rubin Observatory.