SEDMORPH · The origins of galaxy bimodality: Linking mergers, starbursts and feedback in observations and simulations

Understanding how and why galaxies form and evolve is one of the most challenging problems in modern astrophysics. Our own galaxy, the Milky Way, shows order and structure, as do most massive galaxies in our local neighbourhood. Yet when we look to very distant galaxies they are disordered and chaotic. The leading theory for the origin of this transformation invokes gas-rich mergers, which trigger massive starbursts leading to bulge and supermassive black hole growth. The aim of this project is to provide conclusive observational evidence to confirm or refute this fundamental theory of galaxy evolution.Considerable quantities of high quality data are now available for both local and distant galaxies; new methodology is urgently required to enable the translation of this data into an improved understanding of galaxy formation. In this project I will lead a team to develop a suite of new techniques to: (1) statistically link galaxy populations traditionally studied in isolation (starbursts, post-starbursts, mergers, remnants); (2) combine information from both the multi-wavelength spectral energy distributions and morphologies of galaxy samples; (3) visualise the information contained in multiple large datasets. My team will compare directly with merger models to interpret the data in terms of the physical processes driving galaxy evolution. The new techniques will provide stringent observational constraints on models, improve robustness of model-data comparison and highlight areas for improvement.As the only researcher with access to all four of the newest world-leading surveys for galaxy evolution, I am uniquely placed to build an integrated picture of the dominant physical processes that drive galaxy evolution over 3/4 of cosmic time. An ERC grant will allow me to build a team to fully exploit the information provided by all four surveys, through novel analysis techniques and concurrent comparison with models.