The Three Hundred Project: A fast semi-analytic model emulator of hydrodynamical galaxy cluster simulations

Next-generation photometric and spectroscopic surveys will detect faint galaxies in massive clusters, advancing our understanding of galaxy formation in dense environments. Comparing these observations with theoretical models requires high-resolution cluster simulations. Hydrodynamical simulations effectively resolve galaxy properties in halos; however, they face challenges in simulating low-mass galaxies within massive clusters due to computational limitations. On the other hand, dark matter-only (DMO) simulations can provide higher resolution but need models to populate subhalos with galaxies. In this work, we introduce a fast and efficient emulator of hydrodynamical simulations of galaxy clusters, based on the semi-analytic models (SAMs) SAGE and SAG. The calibration of the cluster galaxy properties in the SAMs was guided by the cluster galaxies from the hydrodynamical simulations at intermediate resolution, which represents the highest resolution achievable with current hydrodynamical simulations, ensuring consistency in properties such as stellar masses and luminosities across different redshifts. These SAMs are then applied to DMO simulations from T HE T HREE H UNDRED Project at three different resolutions. Our results show that the SAG model, unlike SAGE, more efficiently emulates the galaxy properties tested in this study even at the highest resolution. This improvement results from the detailed treatment of orphan galaxies, which are satellite galaxies that contribute significantly to the overall galaxy population. SAG enables the study of dwarf galaxies down to stellar masses of M * = 10 7 M ⊙ at the highest resolution, which is an order of magnitude smaller than the stellar masses of galaxies in the hydrodynamical simulations at the intermediate resolution, corresponding to approximately four magnitudes fainter. This demonstrates that a SAM can be effectively calibrated to provide fast and accurate predictions for specific hydrodynamical simulations, offering a computationally efficient alternative for exploring galaxy populations in dense environments across higher resolutions.