Quantifying the impact of AGN feedback on the large-scale matter distribution using two- and three-point statistics

TL;DR

This study uses hydrodynamical simulations to quantify how AGN feedback influences large-scale matter distribution, revealing significant effects on small scales and identifying key epochs where feedback becomes dominant.

Contribution

It demonstrates the impact of varying AGN feedback parameters on matter clustering using 2- and 3-point statistics in EAGLE simulations, highlighting the importance of feedback modeling.

Findings

Varying viscosity has a minor impact (~10%) on large scales.

Changes in AGN heating temperature cause up to 70% variation in small-scale gas clustering.

The redshift range 1.5-1 is critical for AGN feedback influence.

Abstract

Feedback from active galactic nuclei (AGN) plays a critical role in shaping the matter distribution on scales comparable to and larger than individual galaxies. Upcoming surveys such as $\textit{Euclid}$ and LSST aim to precisely quantify the matter distribution on cosmological scales, making a detailed understanding of AGN feedback effects essential. Hydrodynamical simulations provide an informative framework for studying these effects, in particular by allowing us to vary the parameters that determine the strength of these feedback processes and, consequently, to predict their corresponding impact on the large-scale matter distribution. We use the EAGLE simulations to explore how changes in subgrid viscosity and AGN heating temperature affect the matter distribution, quantified via 2- and 3-point correlation functions, as well as higher order cumulants of the matter distribution. We find that varying viscosity has a small impact ($\approx 10\%$) on scales larger than $1 h^{-1}$ Mpc, while changes to the AGN heating temperature lead to substantial differences, with up to $70\%$ variation in gas clustering on small scales ($\lesssim 1 h^{-1}$ Mpc). By examining the suppression of the power spectrum as a function of time, we identify the redshift range $z = 1.5 - 1$ as a key epoch where AGN feedback begins to dominate in these simulations. The 3-point function provides complementary insight to the more familiar 2-point statistics, and shows more pronounced variations between models on the scale of individual haloes. On the other hand, we find that effects on even larger scales are largely comparable.