Priors on Lagrangian bias parameters from galaxy formation modelling

TL;DR

This study evaluates the hybrid Lagrangian bias model's effectiveness in fitting galaxy and halo power spectra across diverse samples, revealing systematic shifts in bias relations and providing priors for future cosmological analyses.

Contribution

It offers a comprehensive analysis of bias parameter relations across various galaxy samples, improving the understanding of galaxy formation effects on bias modeling.

Findings

Hybrid Lagrangian bias model fits all samples accurately.

Bias relations show systematic shifts and larger scatter than halo-only models.

Bias parameter relations can serve as priors in Bayesian cosmological analyses.

Abstract

We study the relations among the parameters of the hybrid Lagrangian bias expansion model, fitting biased auto and cross power spectra up to $k_{\rm max} = 0.7 \, h \, \mathrm{Mpc}^{-1}$. We consider $\sim 8000$ halo and galaxy samples, with different halo masses, redshifts, galaxy number densities, and varying the parameters of the galaxy formation model. Galaxy samples are obtained through state-of-the-art extended subhalo abundance matching techniques and include both stellar-mass and star-formation-rate selected galaxies. All of these synthetic galaxies samples are publicly available at https://bacco.dipc.org/galpk.html. We find that the hybrid Lagrangian bias model provides accurate fits to all of our halo and galaxy samples. The coevolution relations between galaxy bias parameters, although roughly compatible with those obtained for haloes, show systematic shifts and larger scatter. We explore possible sources of this difference in terms of dependence on halo occupation and assembly bias of each sample. The bias parameter relations displayed in this work can be used as a prior for future Bayesian analyses employing the hybrid Lagrangian bias expansion model.