Empirical Dust Attenuation Model Leads to More Realistic UVJ Diagram for TNG100 Galaxies

TL;DR

This paper introduces an empirical dust attenuation model that improves the realism of UVJ diagrams for TNG100 galaxies, aligning better with observations than previous radiative transfer methods.

Contribution

The paper applies a Bayesian dust attenuation model to TNG100 galaxies, producing more accurate UVJ diagrams that match observations better than earlier approximate radiative transfer approaches.

Findings

Empirical model yields UVJ diagrams closer to observed galaxy distributions.

Dust-free UVJ diagram shows qualitative agreement with observations, with minor residual differences.

TNG galaxies are younger and more metal-rich than observed galaxies, affecting the UVJ comparison.

Abstract

Dust attenuation varies substantially from galaxy to galaxy and as of yet cannot be reproduced from first principles in theoretical models. In Nagaraj et al. (2022), we developed the first Bayesian population model of dust attenuation as a function of stellar population properties and projected galaxy shape, built on spectral energy distribution (SED) fits of nearly 30,000 galaxies in the 3D-HST grism survey with broadband photometric coverage from the rest-frame UV to IR. In this paper, we apply the model to galaxies from the large-volume cosmological simulation TNG100. We produce a UVJ diagram and compare it with one obtained in previous work by applying approximate radiative transfer to the simulated galaxies. We find that the UVJ diagram based on our empirical model is in better agreement with observations than the previous effort, especially in the number density of dusty star forming galaxies. We also construct the intrinsic dust-free UVJ diagram for TNG and 3D-HST galaxies at z ~ 1, finding qualitative agreement but residual differences at the 10-20% level. These differences can be largely attributed to the finding that TNG galaxies have, on average, 29% younger stellar populations and 0.28 dex higher metallicities than observed galaxies.