Optimization and Quality Assessment of Baryon Pasting for Intracluster Gas using the Borg Cube Simulation

Abstract: Synthetic datasets generated from large-volume gravity-only simulations are an important tool in the calibration of cosmological analyses. Their creation often requires accurate inference of baryonic observables from the dark matter field. We explore the effectiveness of a baryon pasting algorithm in providing precise estimations of three-dimensional gas thermodynamic properties based on gravity-only simulations. We use the Borg Cube, a pair of simulations originating from identical initial conditions, with one run evolved as a gravity-only simulation, and the other incorporating non-radiative hydrodynamics. Matching halos in both simulations enables comparisons of gas properties on an individual halo basis. This comparative analysis allows us to fit for the model parameters that yield the closest agreement between the gas properties in both runs. To capture the redshift evolution of these parameters, we perform the analysis at five distinct redshift steps, spanning from $z=0$ to $2$. We find that the investigated algorithm, utilizing information solely from the gravity-only simulation, achieves few-percent accuracy in reproducing the median intracluster gas pressure and density, albeit with a scatter of approximately 20%, for cluster-scale objects up to $z=2$. We measure the scaling relation between integrated Compton parameter and cluster mass ($Y_{500c} | M_{500c}$), and find that the imprecision of baryon pasting adds less than 5% to the intrinsic scatter measured in the hydrodynamic simulation. We provide best-fitting values and their redshift evolution, and discuss future investigations that will be undertaken to extend this work.