Title: Water emission tracing active star formation from the Milky Way to high-$z$ galaxies

Abstract: (Abridged) The question of how most stars in the Universe form remains open. While star formation predominantly occurs in young massive clusters, the current framework focuses on isolated star formation. One way to access the bulk of protostellar activity within star-forming clusters is to trace signposts of active star formation with emission from molecular outflows. These outflows are bright in water emission, providing a direct observational link between nearby and distant galaxies. We propose to utilize the knowledge of local star formation as seen with molecular tracers to explore the nature of star formation in the Universe. We present a large-scale statistical galactic model of emission from galactic active star-forming regions. Our model is built on observations of well-resolved nearby clusters. By simulating emission from molecular outflows, which is known to scale with mass, we create a proxy that can be used to predict the emission from clustered star formation at galactic scales. We evaluated the impact of the most important global-star formation parameters (i.e., initial stellar mass function (IMF), molecular cloud mass distribution, star formation efficiency (SFE), and free-fall time efficiency) on simulation results. We observe that for emission from the para-H2O 202 - 111 line, the IMF and molecular cloud mass distribution have a negligible impact on the emission, both locally and globally, whereas the opposite holds for the SFE and free-fall time efficiency. Moreover, this water transition proves to be a low-contrast tracer of star formation. The fine-tuning of the model and adaptation to morphologies of distant galaxies should result in realistic predictions of observed molecular emission and make the galaxy-in-a-box model a tool to analyze and better understand star formation throughout cosmological times.