Enhanced sub-resolution star formation models in cosmological simulations

One of the crucial components in simulating the growth and evolution of galaxies within a cosmological framework is the modeling of star formation (SF) and its corresponding feedback. Traditionally, the implemented SF law follows the empirical Kennicutt-Schmidt relation, which links the SF rate (SFR) in a gas element to the total gas density. More recently, an even stronger correlation has been observed between the SFR and the amount of molecular hydrogen ($\mathrm{H}_2$). This opens up the question of whether molecular hydrogen is a necessary precursor for SF or is instead a tracer of the total amount of gas, both of which would explain the observed correlations. In this study, we examine the impact of using an $\mathrm{H}_2$-based SF law on the formation of Milky Way-mass galaxies using cosmological, hydrodynamical simulations. We find that the galaxy modeled with our approach exhibits a well-defined, disc-like morphology. Compared to the traditional recipe, our model delays the onset of SF by approximately $500 \, \mathrm{Myr}$ resulting in a lower SFR, a smaller disc size, and a higher proportion of neutral to ionized gas within the disc region. These findings highlight the importance of including sophisticated SF models which can be compared to several observations -- including those related to $\mathrm{H}_2$ -- to better understand the processes affecting galaxy formation and evolution.