A synthetic view of HI: mock observations of simulated star-forming galaxies

This thesis presents an analysis of mock H I observations of Milky Way-like galaxies generated from galaxy formation simulations. In particular, the simulations were performed with the state-of-the-art interstellar medium and stellar feedback SMUGGLE model within the moving mesh code AREPO. Mock 21 cm H I emission cubes based on these simulations were built using the MARTINI software, allowing us to investigate the morphology and kinematics of neutral hydrogen across a range of galaxy inclinations. To analyze these mock data cubes, we applied a series of data processing techniques to extract key physical properties of the gas, including moment maps, velocity channel maps, position–velocity diagrams, column density maps, rotation curves, and total H I mass. From these diagnostics, we inferred the structural and kinematic characteristics of the simulated galaxies. We also applied a simple and observationally motivated decomposition method to separate the cold disc from gas located outside the disc—commonly referred to as extra-planar gas (EPG) —that shows anomalous kinematics relative to the gas in the disc. In particular, observational evidence suggests that approximately 10–15% of a galaxy’s H I mass resides in such an extra-planar component (also known as H I halo) and that this gas exhibits a vertical velocity gradient of ∼−10 km s⁻¹ kpc⁻¹ ("lagging halo"). In our analysis, we characterized the kinematics of both the cold disc and the extra-planar gas, identifying their distinct dynamical signatures. We compared our results to observed systems, such as NGC 891 or the HALOGAS survey of nearby galaxies, to assess the realism of the simulated galaxy structures and support the presence of lagging halo gas, finding a qualitative agreement with observations. Our findings underscore the value of synthetic observations as a bridge between simulations and real data, particularly in the context of extra-planar gas dynamics in star-forming galaxies.