Tracing the quenching journey across cosmic time

Abstract

We present the latest version of the GAlaxy Evolution and Assembly (GAEA) theoretical model of galaxy formation. Our new model now combines (i) an updated treatment of feedback from Active Galactic Nuclei (AGN), including an improved modelling of cold gas accretion on super-massive black holes and an explicit implementation of quasar winds; and (ii) a treatment for both cold and hot gas stripping from satellite galaxies. We show that our latest model version predicts specific star formation rate distributions that are in remarkable agreement with observational measurements in the local Universe. Our updated model predicts quenched fractions that are in very nice agreement with observational measurements up to $z 3-4$, and a turn-over of the number densities of quenched galaxies at low stellar masses that is in qualitative agreement with current observational estimates. We show that the main reasons for the improved behaviour with respect to previous renditions of our model are the updated treatment for satellites at low galaxy masses M sun $) and the inclusion of quasar winds at intermediate to large stellar masses ($>10^ M sun $). However, we show that the better treatment of the star formation threshold, due to our explicit partitioning of the cold gas in its atomic and molecular components, also plays an important role in suppressing excessive residual star formation in massive galaxies. While our analysis is based on a selection of quiescent galaxies that takes advantage of the information about their star formation rate, we demonstrate that the impact of a different (colour-colour) selection is not significant up to $z 3$, at least for galaxies above the completeness limits of current surveys. Our new model predicts number densities of massive quiescent galaxies at z > 3 that are the largest among recently published state-of-the-art models. Yet, our model predictions still appear to be below post-JWST observational measurements. We show that the expected cosmic variance is large, and can easily accommodate some of the most recent measurements.