Finding and classifying galaxy mergers
Galaxy mergers underpin our understanding of how galaxies grow and evolve. Our current understanding of the Universe has dark matter halos that grow hierarchically. These dark matter halos typically host a galaxy in their centres, meaning that as the dark matter halos merge, so do the galaxies. The physical processes that go on inside a galaxy, such as star-formation or activate galactic nuclei (AGN) activity, are known to be influenced by a merger event. Merging galaxies are known to show increased star-formation rates (SFR) compared to their non-merging counterparts, and a larger fraction of merging galaxies are found to host AGN compared to non-mergers. The stage of a merger also influences the magnitude of these changes. To study these changes and when they happen, we need to have a way to select merging systems, and identify the merger stage.
This has historically be done by visual inspection, getting people to look at a galaxy and see if it shows signs of interaction. Doing this, however, is time consuming and cannot hope to be scaled to the data volumes of modern surveys, such as Euclid or LSST. It is also unreliable. To solve this, astronomers turned to morphological parameters, with Gini and M20 being two often used statistics to identify galaxy mergers. However, as we will see in this seminar, this method is not as reliable as people often think, only selecting merging galaxies at a certain stage of the merger event.
Selecting the merger stage is also important but difficult. Again, visually we can find galaxies just before a merger (pre-merger) and just after a merger (post-merger) relatively simply (are there two close galaxies or one very messy galaxy?) but this still suffers from a scalability issue. The seminar will also discuss how we can efficiently identify pre- and post-mergers using modern machine learning techniques coupled with simulations.
But we can still take this one step further. The seminar will also explore how we can get the specific time a galaxy is before or after a merger event. Again this will rely on machine learning techniques and some of the latest cosmological simulations. And if time permits, we can look at how the SFR of galaxies changes depending on the time before or after a merger event.