Tracing correlations between galaxy properties across the Cosmic Web: An IllustrisTNG-based study

TL;DR

This study investigates how large-scale cosmic web environments modulate correlations among galaxy properties using the z$=0$ snapshot of the IllustrisTNG TNG300-1 simulation. The authors identify voids, sheets, filaments, and clusters via a Hessian-based deformation tensor, mass-match galaxies across sheet, filament, and cluster to control local density, and quantify both linear and non-linear inter-property relationships with Normalized Mutual Information (NMI), complemented by jackknife error estimates and Student's $t$-tests. They find that correlations among $(u-r)$ color, stellar mass, SFR, and metallicity exist in all environments but are systematically modulated by the web type, with mass–metallicity showing the strongest environmental signal, particularly in filaments, and SFR–metallicity being the weakest; these patterns reflect environment-driven gas accretion, feedback, and interaction processes. The results, while consistent with several observational and simulation studies, are contingent on the specific subgrid physics and the single redshift analyzed, underscoring the need for future work across redshift and with additional properties to robustly map the galaxy–cosmic web connection.

Abstract

We explore the impact of cosmic web environments on galaxy properties such as $(u-r)\,$colour, stellar mass, star formation rate, and stellar metallicity, using a stellar mass-matched sample of simulated galaxies from the IllustrisTNG simulation. We use Normalized Mutual Information (NMI) to quantify correlations among galaxy properties and apply Student's t-test to assess the statistical significance of their differences across cosmic web environments. In every case, the null hypothesis is rejected at $> 99.99\%$ confidence, providing strong evidence that correlations among galaxy properties are strongly dependent on cosmic web environments.

Figures (7)

• Figure 1: This figure shows a projected distribution galaxies in different cosmic web environments within a slice of thickness 10 Mpc in TNG300.
• Figure 2: The left panel and right of this figure shows the PDFs of $\log_{10}\left(\frac{M_{\star}}{M_{\odot}}\right)$, before and after matching the stellar masses of the galaxies, respectively.
• Figure 3: The different panels in this figure illustrate the relationships between various pairs of galaxy properties. The y-axis in each panel represents the median value of a galaxy property, computed with bins of another galaxy property shown along the x axis. Only those bins with at least 10 galaxies are considered for median calculation. The $1\sigma$ errors on each data points are estimated from 50 bootstrap samples drawn from the original dataset.
• Figure 4: The different panels of this figure show the probability distribution functions (PDF) of $\rm (u-r)$ colour, $\rm SFR$, and metallicity. The PDFs in sheets, filaments and clusters are shown together in each panel for comparison.
• Figure 5: This figure illustrates the joint probability distributions of $(u-r)\,$ colour, star formation rate and stellar metallicity with stellar mass of galaxies, in three different types of cosmic web environments.