"Comparing star formation in TNG50 (SH03) with the Pressure-Regulated, Feedback-Modulated (PRFM) Equilibrium Model"

Traditional star formation sub-grid models, such as the Springel and Hernquist (2003, SH03), implemented on cosmological galaxy formation simulations, usually employ phenomenological adjustable parameters to reproduce certain key observables such as the Kennicutt-Schmidt empirical relation. Calibrated to a suite of multiphase magnetohydrodynamic simulations run with the TIGRESS computational framework, Ostriker and Kim (2022) have recently presented the pressure-regulated, feedback-modulated (PRFM) model, which is a more physically motivated model to form stars on kpc-scales, that is based on various local environmental properties such as the gas surface density, midplane stellar density, stellar height, and dark matter volumetric density. We present a detailed comparison between SH03 and PRFM models using outputs from TNG50 in post-processing. We find that the gas height scale is not resolved as compared to theoretical predictions assuming vertical dynamical equilibrium. The PRFM model tends to form stars much more efficiently than SH03 at high gas densities, particularly at high redshift, which can be tested by the upcoming JWST surveys.